Size: 5656
Comment:
|
Size: 103311
Comment:
|
Deletions are marked like this. | Additions are marked like this. |
Line 1: | Line 1: |
= Build Your Own DRMS = The JSOC data archive is designed to be replicable and able to function with other cooperating data archives sharing the same basic architecture. Individual archives can selectively share data in their DRMS databases with other archives and serve as either master or slave for data record information on a dataset by dataset basis. It is also possible for data archives to share cached data segments in their individual Storage Unit Management Systems (SUMS). This page provides information for developers outside the JSOC who wish to set up such cooperating archives. There are three fundamental requiremants for setting up and operating a DRMS system: - A reserved disk space to serve as the SUMS disk cache. (A tape library for permanent offline or near-line storage is nice, but not essential. The details of setting up a tape library are NOT discussed elsewhere.) - A database server running Postgres version ... - A "current" copy of the JSOC software tree, available from Stanford through .... == Setting up a SUMS == The SUMS disk area can be as simple as a directory, but it is probably better to assign at least one disk partition to the SUMS cache. Unless a tape library also exists, the SUMS partition(s) must be large enough to store all the data segments in the DRMS that are to be archived locally. For datasets for which other DRMS servers provide the permanent archive, the local SUMS will serve only as a local cache, so size is dictated by expected usage. The directory or directories to be used for SUMS must be owned by a user named '''production''' (can be any uid) and belong to a group named '''SOI''' (can be any gid), and have a permissions mask of 8354 (''drwxrwsr-x''). The group '''SOI''' should include as members any users who will be writing data into the DRMS by running modules or otherwise. == Setting up the Postgres Database server == You should have Postgres Version 8.1 or higher installed; JSOC database servers are currently (Oct 2006) running on the following systems: * a 64-bit dual-core xeon running Red Hat Enterprise Linux 4 with Postgres v. 8.1.2 * a 32-bit dual-core pentium 4 running Scientific Linux (?; equinox) with Postgres v. 8.1.4 == Populating the Database == First, you must create the database tables required for SUMS. You can do so by running the following psql commands: {{{ create table SUM_MAIN ( ONLINE_LOC VARCHAR(80) NOT NULL, ONLINE_STATUS VARCHAR(5), ARCHIVE_STATUS VARCHAR(5), OFFSITE_ACK VARCHAR(5), HISTORY_COMMENT VARCHAR(80), OWNING_SERIES VARCHAR(80), STORAGE_GROUP integer, STORAGE_SET integer, BYTES bigint, DS_INDEX bigint, CREATE_SUMID bigint NOT NULL, CREAT_DATE timestamp(0), ACCESS_DATE timestamp(0), USERNAME VARCHAR(10), ARCH_TAPE VARCHAR(20), ARCH_TAPE_POS VARCHAR(15), ARCH_TAPE_FN integer, ARCH_TAPE_DATE timestamp(0), WARNINGS VARCHAR(260), STATUS integer, SAFE_TAPE VARCHAR(20), SAFE_TAPE_POS VARCHAR(15), SAFE_TAPE_FN integer, SAFE_TAPE_DATE timestamp(0), constraint pk_summain primary key (DS_INDEX) ); create table SUM_OPEN ( SUMID bigint not null, OPEN_DATE timestamp(0), constraint pk_sumopen primary key (SUMID) ); create table SUM_PARTN_ALLOC ( wd VARCHAR(80) not null, sumid bigint not null, status integer not null, bytes bigint, effective_date VARCHAR(20), archive_substatus integer, group_id integer, ds_index bigint not null, safe_id integer ); create table SUM_PARTN_AVAIL ( partn_name VARCHAR(80) not null, total_bytes bigint not null, avail_bytes bigint not null, pds_set_num integer not null, constraint pk_sumpartnavail primary key (partn_name) ); create table SUM_TAPE ( tapeid varchar(20) not null, nxtwrtfn integer not null, spare integer not null, group_id integer not null, avail_blocks bigint not null, closed integer not null, last_write timestamp(0), constraint pk_tape primary key (tapeid) ); create sequence SUM_SEQ increment 1 start 2 no maxvalue no cycle cache 50; create sequence SUM_DS_INDEX_SEQ increment 1 start 1 no maxvalue no cycle cache 10; create table SUM_FILE ( tapeid varchar(20) not null, filenum integer not null, gtarblock integer, md5cksum varchar(36) not null, constraint pk_file primary key (tapeid, filenum) ); create table SUM_GROUP ( group_id integer not null, retain_days integer not null, effective_date VARCHAR(20), constraint pk_group primary key (group_id) ); }}} (These are contained in the scripts '''create_tables.sql''', '''sum_file.sql''', and '''sum_group.sql''' in the JSOC software library '''base/sums/scripts/postgres'''.) For example, if you have created a database named ''mydb'' on a server named ''myserver'' (and had one of those scripts in your ''wd''), you could enter the command {{{ psql -h myserver mydb -f create_tables.sql }}} Or you could simply enter the commands by hand. (You should be the database administrator when you create these tables.) |
= NetDRMS - a shared data management system = == Introduction == In order to process, archive, and distribute the substantial quantity of solar data captured by the Atmospheric Imaging Assembly (AIA) and Helioseismic and Magnetic Imager (HMI) instruments on the Solar Dynamics Observatory (SDO), the Joint Science Operations Center (JSOC) has developed its own data-management system, NetDRMS. This system comprises two PostgreSQL databases, multiple file systems, a tape back-up system, and software to manage these components. Related sets of data are grouped into data series, each, conceptually, a table of data where each row of data typically associated with an observation time, or a Carrington rotation. As an example, the data series hmi.M_45s contains the HMI 45-second cadence magnetograms, both observation metadata and image FITS files. The columns contain metadata, such as the observation time, the ID of the camera used to acquire the data, the image rotation, etc. One column in this table contains an ID that refers to a set of data files, typically a set of FITS files that contain images. The Data Record Management System (DRMS) is the subsystem that contains and manages the "DRMS" database of metadata and data-file-locator information. One component is a software library, written in C, that provides client programs, also known as "DRMS modules", with an Application Programming Interface (API) that allows the users to access these data. The Storage Unit Management System (SUMS) is the subsystem that contains and manages the "SUMS" database and associated storage hardware. The database contains information needed to locate data files that reside on hardware. The entire system as a whole is typically referred to as DRMS. The user interfaces with the DRMS subsystem only, and the DRMS subsystem interfaces with SUMS - the user does not interact with SUMS directly. The JSOC provides NetDRMS to non-JSOC institutions so that those sites can take advantage of the JSOC-developed software to manage large amounts of solar data. A NetDRMS site is an institution with a local NetDRMS installation. It does not generate the JSOC-owned production data series (e.g., hmi.M_720s, aia.lev1) that Stanford generates for scientific use. A NetDRMS site can generate its own data, production or otherwise. That site can create software that uses NetDRMS to generate its own data series. But it can also act as a "mirror" for individual data series. When acting as a mirror for a Stanford data series, the site downloads from Stanford DRMS database information and stores it in its own NetDRMS database, and it downloads SUMS files, and stores them in its own SUMS subsystem. As the data files are downloaded to the local SUMS, the SUMS database is updated with the information needed to manage the data files. It is possible for a NetDRMS site to mirror the DRMS data of any other NetDRMS site, but at this point, the only site whose data are currently mirrored is the Stanford JSOC. == Installing NetDRMS == Installing the NetDRMS system requires: * installing PostgreSQL [ [[#install-pg|Installing PostgreSQL]] ] * instantiating a PostgreSQL cluster for two databases (one for DRMS and one for SUMS) [ [[#initialize-pg|Initializing PostgreSQL]] ] * installing CFITSIO [ [[#install-cfitsio|Installing CFITSIO] ]] * installing the DBD::Pg Perl package [ [[#install-perl-dbdpg|Installing DBD::Pg]] ] * installing packages to the system Python 3, or installing a new distribution, like Anaconda [ [[#install-python3|Installing Python3]] ] * installing {{{openssl}}} development packages [ [[#install-openssldev|Installing OpenSSL Development Packages]] ] * installing the NetDRMS software code tree, which includes code to create DRMS libraries and modules and SUMS libraries [ [[#install-netdrms|Installing NetDRMS]] ] * initializing SUMS storage such as hard drives or SSD drives [ [[#initialize-sums-disk|Initializing SUMS Storage]] ] * creating DRMS user accounts [ [[#create-users|Creating DRMS User Accounts]]] * running the SUMS daemon (which accepts and processes SUMS requests from DRMS clients) [ [[#run-sums|Running SUMS]]] Optional steps include: * registering for JSOC-data-series subscriptions and running NetDRMS software to receive, in real time, data updates [ [[#register-subscriptions|Registering for Subscriptions]] ] * running the Remote SUMS daemon (which accepts and processes requests for SUs that reside at other NetDRMSs) [ [[#run-remote-sums|Running Remote SUMS]] ] * installing JSOC-specific project code that is not part of the base NetDRMS installation; the JSOC maintains code to generate JSOC-owned data that is not generally of interest to NetDRMS sites, but sites are welcome to obtain downloads of that code. Doing so involves additional configuration to the base NetDRMS system. * installing Slony PostgreSQL data-replication software to become a provider of your site's data * installing a webserver that hosts several NetDRMS CGIs to allow web access to your data * installing the Virtual Solar Observatory (VSO) software to become a VSO provider of data For best results, and to facilitate debugging issues, please follow these steps in order. == Conventions == In this document, parameters to be determined by you the NetDRMS administrator are denoted with angled brackes, {{{<a parameter>}}}. For example, you will need to select a machine to host the PostgreSQL database system, and the name of that host is represented by {{{<PostgreSQL host>}}}. If you choose a host named {{{netdrms_db}}}, let's say, then functionally, you can substitute {{{netdrms_db}}} for {{{<PostgreSQL host>}}} throughout this document. Part of the process of installing NetDRMS is creating a configuration file named {{{config.local}}}. That file contains numerous parameters, such as {{{SUMS_USEMTSUMS}}}. Through this document, those parameters are denoted with square brackets. As such, this document refers to the the parameter {{{SUMS_USEMTSUMS}}} as {{{[SUMS_USEMTSUMS]}}}. <<Anchor(install-pg)>> == Installing PostgreSQL == PostgreSQL is a relational database management system. Data are stored primarily in relations (tables) of records that can be ''mapped'' to each other - given one or more records, you can query the database to find other records. These relations are organized on disk in a hierarchical fashion. At the top level are one or more database ''clusters''. A cluster is simply a storage location on disk (i.e., directory). PostgreSQL manages the cluster's data files with a single process, or PostgreSQL ''instance''. Various operations on the cluster will result in PostgreSQL forking new ephemeral child processes, but ultimately there is only one master/parent process per cluster. Each cluster contains the data for one or more databases. Each cluster requires a fair amount of system memory, so it makes sense to install a single cluster on a single host. It does ''not'' make sense to make separate clusters, each holding one database; each cluster can efficiently support many databases, which are then fairly independent of each other. ''In terms of querying'' the databases are completely independent (i.e., a query on one database cannot involve relations in different databases). However, two databases in a single cluster ''do'' share the same disk directory, so there is not the same degree of independence at the OS/filesystem level. This may only matter if an administrator is operating directly on the files (performing backups, replication, creating standby systems, etc.). To install PostgreSQL, select a host machine, {{{<PostgreSQL host>}}}, to act as the PostgreSQL database server. We recommend installing ''only'' PostgreSQL on this machine, given the large amount of memory and resources required for optimal PostgreSQL operation. We find a Fedora-based system, such as CentOS, to be a good choice, but please visit [[https://www.postgresql.org/docs]] for system requirements and other information germane to installation. The following instructions assume a Fedora-based Linux system such as CentOS (documentation for other distributions, such as Debian and openSUSE can be found online) and a bash shell.<<BR>><<BR>> Install the needed PostgreSQL server packages on {{{<PostgreSQL host>}}} by first visiting [[https://yum.postgresql.org/repopackages.php]] to locate and download the PostgreSQL "repo" rpm file appropriate for your OS and architecture. Each repo rpm contains a {{{yum}}} configuration file that can be used to install all supported PostgreSQL releases. You should install the latest version if possible (version 12, as of the time of this writing). Although you can use your browser to download the file, it might be easier to use Linux command-line tools: {{{ $ curl -OL https://download.postgresql.org/pub/repos/yum/reporpms/EL-7-x86_64/pgdg-redhat-repo-latest.noarch.rpm }}} Install the yum repo configuration file ({{{pgdg-redhat-all.repo}}}) from the downloaded repo rpm file: {{{ $ sudo rpm -i pgdg-redhat-repo-latest.noarch.rpm }}} This installs the repo configuration file to {{{/etc/yum.repos.d/}}}. Find the names of the PostgreSQL packages needed from the repository; the following assumes PostgreSQL 12, but should you want to install an older version, replace "12" with one of 94, 95, 96, 10, or 11: {{{ $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql[0-9]*\.' | cut -d '.' -f 1 postgresql12 $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*devel\.' | cut -d '.' -f 1 postgresql12-devel $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*contrib\.' | cut -d '.' -f 1 postgresql12-contrib $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*libs\.' | cut -d '.' -f 1 postgresql12-libs $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*plperl\.' | cut -d '.' -f 1 postgresql12-plperl $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*server\.' | cut -d '.' -f 1 postgresql12-server }}} Use {{{yum}}} to install all four packages: {{{ $ sudo yum install <packages> }}} where {{{<packages>}}} are the package names determined in the previous step ({{{postgresql12 postgresql12-contrib postgresql12-devel postgresql12-libs postgresql12-plperl postgresql12-server}}}). The rpm package installation will have created the PostgreSQL superuser Linux account {{{<PostgreSQL superuser>}}} (i.e., {{{postgres}}}); {{{<PostgreSQL superuser>}}} will own the PostgreSQL database clusters and server processes that will be created in the following steps. To perform the next steps, you will need to become user {{{<PostgreSQL superuser>}}}: {{{ $ sudo su - <PosgreSQL superuser> }}} Depending on where the package files are installed, you might need to add the PostgreSQL command to your {{{PATH}}} environment variable. To test this, run: {{{ $ which initdb }}} If the {{{initdb}}} command cannot be found, then add the PostgreSQL binaries path to {{{PATH}}}. Find the path to the PostgreSQL installation: {{{ $ rpm -ql postgresql12 <PostgreSQL install dir>/bin/clusterdb ... }}} In this example, {{{<PostgreSQL install dir>}}} is {{{/usr/pgsql-12}}}. Then add the binary path to {{{PATH}}}: {{{ $ export PATH=/usr/pgsql-12/bin:$PATH }}} {{{<PosgreSQL superuser>}}} will be using the binaries in this directory, so it is a good idea to add the export command to {{{.bashrc}}}. As described above, create one database cluster for the two databases (one for DRMS data, and one for SUMS data): {{{ $ whoami postgres $ initdb --locale=C -D <PostgreSQL cluster> }}} where {{{<PostgreSQL cluster>}}} should be {{{/var/lib/pgsql/netdrms}}}. Use this path, unless there is some good reason you cannot. {{{initdb}}} will initialize the cluster data directory (identified by the {{{-D}}} argument). This will result in the creation of template databases, configuration files, and other items.<<BR>><<BR>> The database cluster will contain two configuration files you need to edit: {{{postgresql.conf}}} and {{{pghba.conf}}}. Please refer to the PostgreSQL documentation to properly edit these files. Here are some brief suggestions: * {{{postgresql.conf}}} - for changes to take effect of parameters marked with a ^*^, a restart of a running server instance is required ({{{pg_ctl restart}}}), otherwise changes will require a reload ({{{pg_ctl reload}}}) * {{{listen_addresses}}}^*^ specifies the interface on which the {{{postgres}}} server processes will listen for incoming connections. You will need to ensure that connections can be made from all machines that will run DRMS modules (the modules connect to both the DRMS and SUMS databases), so change the default {{{'localhost'}}} to {{{'*'}}}, which causes the servers to listen on all interfaces: {{{ listen_addresses = '*' }}} * {{{port}}}^*^ is the server port on which the server listens for connections. {{{ port = <PostgreSQL port> }}} The default port is {{{5432}}}, and unless there is a good reason use {{{5432}}} for {{{<PostgreSQL port>}}} * {{{logging_collector}}}^*^ controls whether on not stdout and stderr are logged to a file in the database cluster (in the {{{log}}} or {{{pg_log}}} directory, depending on release). By default it is off - set it so {{{on}}} in each cluster. {{{ logging_collector = on }}} * {{{log_rotation_size}}} sets the maximum size, in kilobytes, of a log file. Set this to {{{0}}} to disable rotation, otherwise a new log will be created after the current one grows to some size. {{{ log_rotation_size = 0 }}} * {{{log_rotation_age}}} set the maximum age, in minutes, of a log file. Set this to {{{1d}}} (1 day) so that each day a new log file is created. {{{ log_rotation_age = 1d }}} * {{{log_min_duration_statement}}} is the amount of time, in milliseconds, a query must run before triggering a log entry. Set to this 1000 so that only long-running queries, over a second, will be logged. {{{ log_min_duration_statement = 1000 }}} * {{{shared_buffers}}}^*^ is the size of shared-memory buffers. For a server dedicated to a single database cluster, this should be about 25% of the total memory. {{{ shared_buffers = 32GB }}} * {{{pg_hba.conf}}} controls the methods by which client authentication is achieved (HBA stands for host-based authentication). It will likely take a little time to understand and properly edit this configuration file. If you are not familiar with networking concepts (such as subnets, name resolution, reverse name resolution, CIDR notation, IPv4 versus IPv6, network interfaces, etc.) then now is the time to become familiar.<<BR>><<BR>>This configuration file contains a set of columns that identify which user can access which database from which machines. It also defines the method by which authenticaton occurs. When a user attempts to connect to a database, the server transverses this list looking for the ''first'' row that matches. Once this row is identified, the user must authenticate - if authentication fails, the connection is rejected. The server does '''''not''''' attempt additional rows. For changes to take effect of any of the parameters in this file, a {{{reload}}} of a server instance is required (not a {{{restart}}}) Here are the recommended entries: {{{ # local superuser connections # TYPE DATABASE USER AUTH-METHOD local all all trust # this applies ONLY if the user is logged into the PG server AND they do not use the -h argument to psql host all all 127.0.0.1/8 trust # for -h localhost, if localhost resolves to an IPv4 address; also for -h 127.0.0.1 host all all ::1/128 trust # for -h localhost, if localhost resolves to an IPv6 address; also for -h ::1 # non-local superuser connections # TYPE DATABASE USER ADDRESS AUTH-METHOD host all postgres XXX.XXX.XXX.XXX/YY trust # non-superuser connections (which can be made from any non-server machines only) # TYPE DATABASE USER ADDRESS AUTH-METHOD host netdrms all XXX.XXX.XXX.XXX/YY md5 host netdrms_sums all XXX.XXX.XXX.XXX/YY md5 }}} where the columns are defined as follows: * {{{TYPE}}} - this column defines the type of socket connection made (Unix-domain, TCP/IP, the encryption used, etc.). {{{local}}} is ''only'' relevant to Unix-domain local connections from the database server host {{{<PostgreSQL host>}}} itself. Since only {{{postgres}}} will log into the database server, the first row above applies to the administrator only. {{{host}}} is ''only'' relevant to TCP/IP connections, regardless of the encryption status of the connection. * {{{DATABASE}}} - this column identifies the database to which the user has access. Whenever a user attempt to connect to the database server, they specify a database to access. That database must be in the DATABASE column. We recommend using {{{netdrms}}} for non-superusers, blocking such users from accessing all databases except the DRMS one. Conversely, we recommend using {{{all}}} for the superuser so they can access both the DRMS and SUMS databases (and any other that might exist).<<BR>><<BR>> NOTE: You are using the database name {{{netdrms}}} in {{{pg_hba.conf}}}, even though you have not actually created that database yet. This is OK; you will do so once you start the PostgreSQL cluster instance. * {{{USER}}} - this column identifies which users can access the specified databases. * {{{ADDRESS}}} - this column identifies the host IP addresses (or host names, but do not use those) from which a connection is allowed. To specify a range of IP addresses, such as those on a subnet, use a CIDR address. This column should be empty for {{{local}}} connections. * {{{AUTH-METHOD}}} - this column identifies the type authentication to use. We recommend using either {{{trust}}} or {{{md5}}}. When {{{trust}}} is specified, PostreSQL will unconditionally accept the connection to the database specified in the row. If {{{md5}}} is specified, then the user will be required to provide a password. If you follow the recommendations above, then for the {{{local}}} row, any user who can log into the database server can access any database in the cluster without any further authentication. Generally only a superuser will be able to log into the database server, so this choice makes sense. For non-{{{local}}} connections by {{{postgres}}}, the Linux PostreSQL superuser {{{postgres}}} can access any database on the server without further authentication. For the remaining non-{{{local}}} non-{{{postgres}}} connections, users will need to provide a password. Should you need to edit either of these configuration files AFTER you have started the database instance (by running {{{pg_ctl start}}}, as described in the next section), you will need to either {{{reload}}} or {{{restart}}} the instance: {{{ $ whoami postgres # reload $ pg_ctl reload -D <PostgreSQL cluster> # restart $ pg_ctl restart -D <PostgreSQL cluster> }}} <<Anchor(initialize-pg)>> == Initializing PostgreSQL == You need to now initialize your PostgreSQL instance by creating the DRMS and SUMS databases, installing database-server languages, creating a schema, creating a relation. To accomplish this become {{{<PostgreSQL superuser>}}}; all steps in this section must be performed by the superuser: {{{ $ sudo su - postgres }}} Start the database instance for the cluster you created: {{{ $ whoami postgres $ pg_ctl start -D <PostgreSQL cluster> }}} You previous created {{{<PostgreSQL cluster>}}}, which will most likely be {{{/var/lib/pgsql/netdrms}}}. Ensure the configuration files you created work. This can be done by attempting to connect to the database server as {{{<PostgreSQL superuser>}}} with {{{psql}}} from {{{<PostgreSQL host>}}}: {{{ $ whoami <PostgreSQL superuser> $ hostname <PostgreSQL host> $ psql -h <PostgreSQL host> -p <PostgreSQL port> psql (12.1) Type "help" for help. postgres=# \q $ }}} The PostgreSQL installation resulted in the creation of the {{{postgres}}} database superuser, and since {{{psql}}} connects to the database as the database user with the same name as the Linux user running {{{psql}}}, you will be logged in as database user {{{postgres}}}. This is indicated by the {{{postgres=#}}} prompt (the hash refers to a superuser). After you successfully see the superuser prompt, create the two databases: {{{ $ whoami postgres # create the DRMS database $ createdb --locale C -E UTF8 -T template0 netdrms # create the SUMS database $ createdb --locale C -E UTF8 -T template0 netdrms_sums }}} Install the required database-server languages: {{{ $ whoami postgres # create the PostgreSQL scripting language (versions <= 9.6) # no need to create the PostgreSQL scripting language (versions > 9.6) $ createlang plpgsql netdrms # create the "trusted" perl language (versions <= 9.6) createlang -h <PostgreSQL host> plperl netdrms # create the "trusted" perl language (versions > 9.6) $ psql -h <PostgreSQLhost> -p <PostgreSQL port> netdrms netdrms=# CREATE EXTENSION IF NOT EXISTS plperl; netdrms=# \q # create the "untrused" perl language (versions <= 9.6) $ createlang -h <PostgreSQL host> plperlu netdrms # create the "untrused" perl language (versions > 9.6) netdrms=# CREATE EXTENSION IF NOT EXISTS plperlu; netdrms=# \q }}} The SUMS database does not use any language extensions so there is no need to create any for the SUMS database. At this point, it is a good idea to create a password for the {{{postgres}}} database superuser: {{{ $ whoami <PostgreSQL superuser> $ psql -h <PostgreSQLhost> -p <PostgreSQL port> netdrms netdrms=# ALTER ROLE postgres WITH PASSWORD '<new password>'; ALTER ROLE netdrms=# \q $ }}} <<Anchor(install-cfitsio)>> == Installing CFITSIO == The base NetDRMS release requires CFITSIO, a {{{C}}} library used by NetDRMS to read and write FITS files. Visit [[https://heasarc.gsfc.nasa.gov/fitsio/]] to obtain the link to the CFITSIO source-code tarball. Create an installation directory {{{<CFITSIO install dir>}}} (such as {{{/opt/cfitsio-X.XX}}}, with a link from {{{/opt/cfitsio}}} to {{{<CFITSIO install dir>}}}), download the tarball, and extract the tarball into {{{<CFITSIO install dir>}}}: {{{ $ sudo mkdir -p <CFITSIO install dir> $ ctrl+d }}} {{{ $ curl -OL 'http://heasarc.gsfc.nasa.gov/FTP/software/fitsio/c/cfitsio-X.XX.tar.gz' $ tar xvzf netdrms_X.X.tar.gz $ cd cfitsio-X.XX }}} Please read the {{{README}}} file for complete installation instructions. As a quick start, run: {{{ $ ./configure --prefix=<CFITSIO install dir> # build the CFITSIO library $ make # install the libraries and binaries to <CFITSIO install dir> $ sudo make install # create the link from cfitsio to <CFITSIO install dir> $ sudo su - $ cd <CFITSIO install dir>/.. $ ln -s <CFITSIO install dir> cfitsio }}} CFITSIO has a dependency on {{{libcurl}}} - in fact, any program made by linking to {{{cfitsio}}} will also require {{{libcurl-devel}}} since {{{cfitsio}}} uses the {{{libcurl}}} API. We recommend using {{{yum}}} to install the two packages (if they are not already installed - it is quite likely that {{{libcurl}}} will already be installed): {{{ $ sudo yum install libcurl-devel }}} <<Anchor(install-openssldev)>> == Installing OpenSSL Development Packages == NetDRMS requires the OpenSSL Developer's API. If this API has not already been installed, do so now: {{{ $ sudo yum install openssl-devel }}} <<Anchor(install-perl-dbdpg)>> == Installing DBD::Pg == One step in the installation process will require running a {{{perl}}} script that accesses the PostgreSQL database. In order for this to work, you will need to ensure the {{{DBD::Pg}}} module has been installed. To check for installation, run: {{{ $ perl -M'DBD::Pg' }}} If there is no error about not being able to locate the module, and the command simply hangs, then you are all set (enter {{{ctrl-C}}} to exit). If the module is not installed, and you are running the {{{perl}}} installed with the system, then run {{{yum}}} to identify the package: {{{ $ yum list | grep -i 'dbd-pg' ... perl-DBD-Pg.x86_64 2.19.3-4.el7 base ... }}} then bringing to bear all your powers of divination, choose the correct package and install it: {{{ $ sudo yum install 'perl-DBD-Pg' }}} If using a non-system {{{perl}}}, use the distro's installation method. If the distro does not have that module, or the distro installer does not work, as a final act of desperation use CPAN {{{ $ sudo perl -MCPAN -e 'install DBD::Pg' }}} <<Anchor(install-python3)>> == Installing Python3 == NetDRMS requires that a number of {{{python}}} packages and modules be present that are not generally part of a system installation. In addition, many scripts require {{{python3}}} and not {{{python2}}}. The easiest way to satisfy these eeds is to install a ''data-science''-oriented {{{python3}}} distribution, such as {{{Anaconda}}}. In that vein, install {{{Anaconda}}} into an appropriate installation directory such as {{{/opt/anaconda3}}}. To locate the {{{Linux}}} installer, visit [[https://docs.anaconda.com/anaconda/install/linux/]]: {{{ $ curl -OL 'https://repo.anaconda.com/archive/Anaconda3-2019.10-Linux-x86_64.sh' $ sha256sum Anaconda3-2019.10-Linux-x86_64.sh 46d762284d252e51cd58a8ca6c8adc9da2eadc82c342927b2f66ed011d1d8b53 Anaconda3-2019.10-Linux-x86_64.sh $ sudo bash Anaconda3-2019.10-Linux-x86_64.sh }}} After some initial prompts, the installer will display {{{ PREFIX=/home/<user>/anaconda3 }}} This path is the default installation directory ({{{<user>}}} is the user running {{{bash}}}). Replace the {{{PREFIX}}} path with {{{<Anaconda3 install dir>}}}. <<Anchor(install-netdrms)>> == Installing NetDRMS == To install NetDRMS, you will need to select an appropriate machine on which to install NetDRMS, an appropriate machine/hardware on which to host the SUMS service, create Linux users and groups, download the NetDRMS release tarball and extract the release source, initialize the Linux environment, create log directories, create the configuration file and run the configuration script, compile and install the executables, create the the DRMS- and SUMS-database users/relations/functions/objects, initialize the SUMS storage hardware, install the SUMS and Remote SUMS daemons. The optimal hardware configuration will likely depend on your needs, but the following recommendations should suffice for most sites. DRMS and SUMS can share a single host machine. The most widely used and tested Linux distributions are Fedora-based, and at the time of this writing, CentOS is the most popular. Sites have successfully used openSUSE too, but if possible, we would recommend using CentOS. SUMS requires a large amount of storage to hold the DRMS data-series data/image files. The amount needed can vary widely, and depends directly on the amount of data you wish to keep online at any given time. Most NetDRMS sites mirror some amount of (but not all) JSOC SDO data - the more data mirrored, the larger the amount of storage needed. To complicate matters, a site can also mirror only a subset of each data series' data; perhaps one site wishes to retain only the current month's data of many data series, but another wishes to retain all data for one or two series. To decide on the amount of storage needed, you will have to ask the JSOC how much data each series comprises and decide how much of that data you want to keep online. Data that goes offline can always be retrieved automatically from the JSOC again. Data will arrive each day, so request from the JSOC an estimate of the rate of data growth. We recommend doing a rough calculation based upon these considerations, and then doubling the resulting number and installing that amount of storage. Next, create a production Linux user {{{<NetDRMS production user>}}} (named {{{netdrms_production}}} by default): {{{ $ sudo useradd <NetDRMS production user> $ sudo passwd <NetDRMS production user> Changing password for user <NetDRMS production user>. New password: Retype new password: passwd: all authentication tokens updated successfully. $ }}} NOTE: ensure that {{{<NetDRMS production user>}}} is a valid PostgreSQL ''name'' because NetDRMS makes use of the PostgreSQL feature whereby attempts to connect to a database are made as the database user whose name matches the name of the Linux user connecting to the database. Please see [[https://www.postgresql.org/docs/12/sql-syntax-lexical.html#SQL-SYNTAX-IDENTIFIERS]] for a description of valid PostgreSQL names. NetDRMS requires additional Python packages not included in the Anaconda distribution, but if you install Anacoda, then the number of additional packages you need to install is minimal. If you have a different {{{python}}} distribution, then you may need to install additional packages. To install new {{{Anaconda}}} packages, as {{{<NetDRMS production user>}}} first create a virtual environment for NetDRMS (named {{{netdrms}}}): {{{ $ whoami <NetDRMS production user> # Sets path to Anaconda3 binaries so that the next conda command will success (will edit .bashrc) $ <Anaconda install dir>/bin/conda init bash $ source ~/.bashrc $ conda create --name netdrms Collecting package metadata (current_repodata.json): done Solving environment: done ## Package Plan ## environment location: /home/netdrms_production/.conda/envs/netdrms Proceed ([y]/n)? y Preparing transaction: done Verifying transaction: done Executing transaction: done # # To activate this environment, use # # $ conda activate netdrms # # To deactivate an active environment, use # # $ conda deactivate }}} VERY IMPORTANT!! As described above in the {{{conda create}}} output, run: {{{ $ conda activate netdrms }}} If you do not do this, then conda might invoke the system Python, which could be version 2.7, when in fact your Anaconda is Python 3. If that happens, then all hope is lost and you'll need to remove the {{{netdrms}}} virtual environment and re-create it. You will use {{{conda}}} to install additional Anaconda3 packages. To succees, you must have already installed PostgreSQL, and the PostgreSQL executables must be in {{{$PATH}}} (one step in the installation process is running {{{pg_config}}}. Add the following to {{{<NetDRMS production user>}}}'s {{{.bashrc}}} ensure PostgreSQL executables can be located: {{{ $ whoami <NetDRMS production user> $ vi ~/.bashrc i ... # PostgreSQL executables export PATH=<PostgreSQL install dir>/bin:$PATH ESC :wq $ }}} Make sure you have either re-logged-in or sourced {{{.bashrc}}}. Then install the new Anaconda packages using {{{conda}}}: {{{ $ whoami <NetDRMS production user> $ conda install -n netdrms psycopg2 psutil python-dateutil Collecting package metadata (current_repodata.json): done ... }}} NOTE: By installing these packages in the {{{netdrms}}} virtual environment, you will also install the {{{python}}} package in the environment since the explicitly listed packages have a dependency on {{{python}}}. This is important, because the next step requires that {{{pip}}}, part of {{{python}}}, be present in the environment. The NetDRMS script needed for series subscription requires installation of the {{{pySmartDL}}} Python 3 package - a package that generally is not part of any distro. Unfortunately, {{{conda}}} supports this package for Python 2.7 only, and NetDRMS requires Python 3. But {{{pip}}} supports a Python 3 version. To install Python 3 {{{pySmartDL}}}, first make sure you have activated your virtual environment, as described above. Then run: {{{ $ whoami <NetDRMS production user> $ which pip <NetDRMS production user home dir>/.conda/envs/netdrms/bin/pip $ pip install pySmartDL }}} If {{{which pip}}} displays an Anaconda path in {{{<Anaconda3 install dir>}}}, then do not run {{{pip}}} just yet. Go back and activate your virtual environment first. From now on, the {{{<NetDRMS production user>}}} should use the virtual environment. You should make sure that {{{$PYTHONPATH}}} is not set, otherwise it might interfere with the running of Anaconda3 {{{python}}}. Modify that user's {{{.bashrc}}} to do so: {{{ $ whoami <NetDRMS production user> $ vi ~/.bashrc i # .bashrc # use Python virtual environment by default unset PYTHONPATH # call conda activate AFTER # <<< conda initialize <<< conda activate netdrms ESC :wq $ }}} For the changes to {{{.bashrc}}} to take effect, either logout/login or source {{{.bashrc}}}. All non-production users will use the {{{netdrms}}} virtual environment. Make sure that users can access the environment: {{{ $ whoami <NetDRMS production user> $ chmod o+x <NetDRMS production user home dir> }}} Create the Linux group {{{<SUMS users>}}}, e.g. {{{sums_users}}}, to which all SUMS users belong, including {{{<NetDRMS production user>}}}. This group will be used to ensure that all SUMS users can create new data files in SUMS: {{{ $ sudo groupadd <SUMS users> }}} Add {{{<NetDRMS production user>}}} to this group (later you will add each SUMS user - users who will read/write SUMS data files - to this group as well): {{{ $ sudo usermod -a -G <SUMS users> <NetDRMS production user> $ id <NetDRMS production user> uid=1001(netdrms_production) gid=1001(netdrms_production) groups=1001(netdrms_production),1002(sums_users) }}} On {{{NetDRMS host}}}, select a path {{{<NetDRMS root>}}} for the root directory of the NetDRMS source tree installed by {{{<NetDRMS production user>}}}. A typical choice for {{{<NetDRMS root>}}} is {{{/opt/netdrms}}} or {{{/usr/local/netdrms}}}, and a typical strategy is to install the source tree into a directory that contains the release version in its name ({{{<NetDRMS install dir>}}}), e.g., {{{/opt/netdrms-9.3}}}. Then {{{<NetDRMS production user>}}} makes a link from {{{<NetDRMS root>}}} to {{{<NetDRMS install dir>}}}. This facilitates the maintenance of multiple releases. To switch between releases, {{{<NetDRMS production user>}}} simply updates the link to point to the desired release directory. Create {{{<NetDRMS install dir>}}} and make {{{<NetDRMS production user>}}} the owner: {{{ $ sudo mkdir -p <NetDRMS install dir> $ sudo chown <NetDRMS production user>:<NetDRMS production user> <NetDRMS install dir> }}} As {{{<NetDRMS production user>}}}, obtain a NetDRMS tarball from [[http://jsoc.stanford.edu/netdrms/dist/]] and extract it into a release-specific directory: {{{ $ whoami <NetDRMS production user> $ cd <NetDRMS install dir> $ curl -OL 'http://jsoc.stanford.edu/netdrms/dist/netdrms_X.X.tar.gz' $ tar xvzf netdrms_X.X.tar.gz $ <ctrl-d> }}} Create the link from <NetDRMS root> to <NetDRMS install dir>: {{{ $ cd <NetDRMS install dir>/.. $ sudo ln -s <NetDRMS install dir> netdrms $ }}} As {{{<NetDRMS production user>}}}, set the two environment variables that are needed for proper NetDRMS operation. To do so, you'll need first to determine the appropriate {{{<architecture>}}} string for one of these variables: {{{ $ whoami <NetDRMS production user> $ cd <NetDRMS root> $ build/jsoc_machine.csh <architecture> }}} It is best to set the following two environment variables in {{{<NetDRMS production user>}}}'s {{{.bashrc}}} file since they must always be set whenever any NetDRMS code is run. And please put the NetDRMS executables in the path: {{{ $ whoami <NetDRMS production user> $ vi ~/.bashrc i # .bashrc export JSOCROOT=/opt/netdrms export JSOC_MACHINE=<architecture> export PATH=/opt/netdrms/bin/<architecture>:$PATH ESC :wq $ }}} Make the SUMS log directory {{{<SUMS logs>}}} on the SUMS server machine. Various SUMS log files will be written to this directory. A suitable directory would reside in the {{{<NetDRMS production user>}}} user's home directory, e.g., {{{$HOME/log/SUMS}}} {{{ $ whoami <NetDRMS production user> $ mkdir -p <SUMS logs> }}} Select appropriate {{{C}}} and {{{Fortran}}} compilers. The DRMS part of NetDRMS must be compiled with a {{{C}}} compiler. NetDRMS supports both the GNU {{{C}}} compiler ({{{gcc}}}), and the Intel {{{C++}}} compiler ({{{icc}}}). Certain JSOC-specific code requires {{{Fortran}}} compilation. For those projects, NetDRMS supports the GNU {{{Fortran}}} compiler ({{{gfortran}}}), and the Intel {{{Fortran}}} compiler ({{{ifort}}}). SUMS is implemented as a Python daemon, so no compilation step is needed. Both GNU and Intel are widely used, so feel free to use either. By default, Intel compilers are used. There are two methods for changing the compilers: * as the {{{<NetDRMS production user>}}}, you can set the {{{JSOC_COMPILER}}} and {{{JSOC_FCOMPILER}}} environment variables (we recommend doing so in {{{.bashrc}}}): {{{ $ whoami <NetDRMS production user> $ vi i # .bashrc # set COMPILER to icc for the Intel C++ compiler, and to gcc for the GNU C++ compiler export JSOC_COMPILER=<C compiler> # set to ifort for the Intel Fortran compiler, and to gfortran for the GNU Fortran compiler export JSOC_FCOMPILER=<Fortran compiler> ESC :wq $ }}} * you can edit the {{{make_basic.mk}}} file in {{{<NetDRMS root>}}}; to select the Intel compilers, edit the {{{COMPILER}}} and {{{FCOMPILER}}} {{{make}}} variables declared near the top of the file: {{{ # use Intel compilers COMPILER = icc FCOMPILER = ifort # use GNU compilers COMPILER = gcc FCOMPILER = gfortran }}} Create the {{{<NetDRMS root>/config.local}}}, the master configuration file for both DRMS and SUMS, using {{{<NetDRMS root>/config.local.newstyle.template}}} as a template. This file contains a number of configuration parameters, along with detailed descriptions of what they control and suggested values for those parameters. The configuration script, {{{configure}}}, reads this file, and then creates one output file, {{{drmsparams.*}}}, in {{{<NetDRMS root>/localization}}} for each of several programming languages/tools ({{{C}}}, {{{GNU make}}}, {{{perl}}}, {{{python}}}, {{{bash}}}). In this manner, the parameters are directly readable by several languages/tools used by NetDRMS. Lines that start with whitespace or the hash symbol, {{{#}}} are ignored. NOTE: if you have an older NetDRMS and you are going to use the {{{config.local}}} from that NetDRMS installation for a new NetDRMS installation, then you might have an ''old-style'' {{{config.local}}}. You know you have an old-style configuration file if the {{{__STYLE__}}} section does not exist in the file, or if it does exist, the value in that section is {{{old}}}. If that is the case, then you will need to compare {{{<NetDRMS root>/config.local.template}}} in the old installation to the analogous file in the new installation to determine the set of parameters that have changed between releases. Several sections compose {{{config.local}}}: {{{ __STYLE__ new __DEFS__ # these are NetDRMS-wide parameter values; the format is <quote code>:<parameter name><whitespace>+<parameter value>; # the configuration script uses <quote code> to assist in creating language-specific parameters; <quote code> is one of: # 'q' (enclose the parameter value in double quotes) # 'p' (enclose the parameter value in parentheses) # 'a' (do not modify the parameter value). __MAKE__ # these are make variables used by the make system during compilation - they generally contain paths to third-party code }}} Before creating {{{config.local}}}, please request from the JSOC a value for {{{DRMS_LOCAL_SITE_CODE}}}. This code uniquely identifies each NetDRMS installation. Each site requires one ID for each of its NetDRMS installations. The {{{__DEFS__}}} section: * {{{BIN_PY3}}} - the path to the Python 3 {{{python}}} executable. * {{{DBNAME}}} - the name of the DRMS database: this is {{{netdrms}}}; this parameter exists in case you want to select a different name, but we don't recommend changing it. * {{{DRMS_LOCAL_SITE_CODE}}} - a 15-bit hexadecimal string that globally and uniquely identifies the NetDRMS. Each NetDRMS requires a unique code for each installation. Values greater than or equal to {{{0x4000}}} denote a development installation and need not be unique. If you plan on generating data that will be distributed outside of your site, please obtain a unique value from the JSOC. * {{{DRMS_LOCK_DIR}}} - the directory to which the DRMS library writes various lock files. * {{{DRMS_LOG_DIR}}} - the directory to which the DRMS library writes various log files. * {{{EXPORT_HANDLE_DIR}}} - the directory to which export programs save handles. * {{{EXPORT_LOCK_DIR}}} - the directory to which export programs write lock files. * {{{EXPORT_LOG_DIR}}} - the directory to which export programs write logs. * {{{JMD_IS_INSTALLED}}} - if set to 1, then the Java Mirroring Daemon alternative to Remote SUMS is used: this should be {{{0}}}. * {{{POSTGRES_ADMIN}}} - the Linux user that owns the PostgreSQL installation and processes: this is {{{<PostgreSQL superuser>}}}. * {{{RS_BINPATH}}} - the NetDRMS binary path that contains the external programs needed by the Remote SUMS (e.g., {{{jsoc_fetch}}}, {{{vso_sum_alloc}}}, {{{vso_sum_put}}}). * {{{RS_DBHOST}}} - the name of the Remote SUMS database cluster host; this is {{{<PostgreSQL host>}}}, the machine on which PostgreSQL was installed. * {{{RS_DBNAME}}} - the Remote SUMS database - this is {{{netdrms_remotesums}}}. * {{{RS_DBPORT}}} - the port that the Remote SUMS database cluster instance is listening on: this is {{{<PostgreSQL port>}}} * {{{RS_DBUSER}}} - the Linux user that runs Remote SUMS; this is also the database user who owns the Remote SUMS database objects: this is {{{<NetDRMS production user>}}}. * {{{RS_DLTIMEOUT}}} - the timeout, in seconds, for an SU to download. If the download time exceeds this value, then all requests waiting for the SU to download will fail. * {{{RS_LOCKFILE}}} - the (advisory) lockfile used by Remote SUMS to prevent multiple instances from running. * {{{RS_LOGDIR}}} - the directory in which remote-sums log files are written. * {{{RS_MAXTHREADS}}} - the maximum number of SUs that Remote SUMS can process simultaneously. * {{{RS_N_WORKERS}}} - the number of {{{scp}}} worker threads - at most, this many {{{scp}}} processes will run simultaneously * {{{RS_REQTIMEOUT}}} - the timeout, in seconds, for a new SU request to be accepted for processing by the daemon. If the daemon encounters a request older than this value, it will reject the new request. * {{{RS_REQUEST_TABLE}}} - the Remote SUMS database relation that contains Remote SUMS requests; this is {{{<Remote SUMS requests>}}}; DRMS modules insert request rows in this table, and Remote SUMS locates the requests and manages rows in this table. * {{{RS_SCP_MAXPAYLOAD}}} - the maximum total payload, in MB, per download. As soon as the combined payload of SUs ready for download exceeds this value, then the SUs are downloaded with a single {{{scp}}} process. * {{{RS_SCP_MAXSUS}}} - the maximum size of the SU download queue. As soon as this many SUs are ready for download, they are downloaded with a single {{{scp}}} process. * {{{RS_SCP_TIMEOUT}}} - if there are SUs ready for download, and no {{{scp}}} has fired off within this many seconds, then the SUs that are ready to download are downloaded with a single {{{scp}}} process. * {{{RS_SITE_INFO_URL}}} - the service at JSOC that is used by Remote SUMS to locate the NetDRMS site that owns SUMS storage units; this is {{{Remote SUMS site URL}}}. * {{{RS_SU_EXPIRATION}}} - the default expiration date for all SUs ingested by Remote SUMS; if the SU being ingested is part of a data series, then Remote SUMS obtains the expiration for the SU from the data series' definition instead; as an alternative to {{{RS_SU_EXPIRATION}}}, {{{RS_SU_LIFESPAN}}} can be used to specify the expiration date of newly ingested SUs; {{{RS_SU_EXPIRATION}}} takes predent over {{{RS_SU_LIFESPAN}}}. * {{{RS_SU_ARCHIVE}}} - the default value of the archive flag for newly ingested SUs; if the SU being ingested is part of a data series, then Remote SUMS obtains the archive flag from the data series' definition instead; the truth value can be one of several character strings that implies TRUE or FALSE. * {{{RS_SU_LIFESPAN}}} - the default lifespan ("retention time"), in days, of a newly ingested SU; if the SU being ingested is part of a data series, then Remote SUMS obtains the lifespan for the SU from the data series' definition instead; as an alternative to {{{RS_SU_LIFESPAN}}}, {{{RS_SU_EXPIRATION}}} can be used to specify the lifespan of newly ingested SUs; {{{RS_SU_EXPIRATION}}} takes predent over {{{RS_SU_LIFESPAN}}}. * {{{RS_SU_TAPEGROUP}}} - the default value of the tapegroup for newly ingested SUs; if the SU being ingested is part of a data series, then Remote SUMS obtains the tapegroup from the data series' definition instead. * {{{RS_TMPDIR}}} - the temporary directory into which SUs are downloaded. This should be on the same file system on which the SUMS partitions reside. * {{{SCRIPTS_EXPORT}}} - the path to the directory in the NetDRMS installation that contains the export scripts. * {{{SERVER}}} - the name of the DRMS database cluster host: this is {{{<PostgreSQL host>}}}, the machine on which PostgreSQL was installed. * {{{SS_HIGH_WATER}}} - partition scrubbing is initiated only after partition percent usage rises above the high-water mark. * {{{SS_LOCKFILE}}} - the (advisory) lockfile used by the SU steward to prevent multiple instances of the steward from running. * {{{SS_LOW_WATER}}} - each SUMS partition is scrubbed until its percent usage falls below the low-water mark. * {{{SS_REHYDRATE_INTERVAL}}} - the time interval, in seconds, between updates to the per-partition cache of expired SUs; this value applies to all partitions that are scrubbed; for each partition, a steward thread queries its cache to select the next SUs to delete (which are sorted by increasing expiration date). * {{{SS_SLEEP_INTERVAL}}} - the interval, in seconds, between flushing/caching expired SU lists (use a smaller number if the system experience a high rate of SU expiration). * {{{SS_SU_CHUNK}}} - the number of SUs in a partition that are deleted at one time; SUs are deleted one chunk at a time until the partition usage falls below the low-water mark. * {{{SUMBIN_BASEDIR}}} - the directory in which sum_chmown, a root setuid program is installed; must be mounted locally on the machine on which the SUMS partition are mounted: this is {{{<NetDRMS root>}}}. * {{{SUMLOG_BASEDIR}}} - the path to the directory that contains various SUMS log files; this is {{{<SUMS logs>}}}. * {{{SUMPGPORT}}} - the port that the SUMS database cluster host is listening on: this is {{{<PostgreSQL port>}}}, unless DRMS and SUMS reside in different clusters on the same host (something that is not recommended since a single PostgreSQL cluster requires a substantial amount of system resources). * {{{SUMS_DB_HOST}}} - the name of the SUMS database cluster host: this is {{{<PostgreSQL host>}}}, the machine on which PostgreSQL was installed; NetDRMS allows for creating a second cluster for SUMS, but in general this will not be necessary unless extremely heavy usage requires separating the two clusters. * {{{SUMS_GROUP}}} - the name of the Linux group to which all SUMS Linux users belong: this is {{{<SUMS users>}}}. * {{{SUMS_MANAGER}}} - the SUMS database user who owns the SUMS database objects which are manipulated by Remote SUMS and SUMS itself; it should be the Linux user that runs SUMS and owns the SUMS storage directories - this is {{{<NetDRMS production user>}}} * {{{SUMS_MULTIPLE_PARTNSETS}}} - SUMS has more than one partition set: more than likely, this is {{{0}}}. * {{{SUMS_READONLY_DB_USER}}} - the SUMS database user who has read-only access to the SUMS database objects; it is used by the Remote SUMS client ({{{rsums-clientd.py}}}) to check for the presence of SUs before requesting they be downloaded. * {{{SUMS_TAPE_AVAILABLE}}} - SUMS has a tape-archive system. * {{{SUMS_USEMTSUMS}}} - use the multi-threaded Python SUMS: this is {{{1}}}. * {{{SUMS_USEMTSUMS_ALL}}} - use the multi-threaded Python SUMS for all SUMS API methods; {{{SUMS_USEMTSUMS_ALLOC}}}, {{{SUMS_USEMTSUMS_CONNECTION}}}, {{{SUMS_USEMTSUMS_DELETESUS}}}, {{{SUMS_USEMTSUMS_GET}}}, {{{SUMS_USEMTSUMS_INFO}}}, and {{{SUMS_USEMTSUMS_PUT}}} are ignored: this is {{{1}}}. * {{{SUMS_USEMTSUMS_ALLOC}}} - use the MT SUMS daemon for the SUM_alloc() and SUM_alloc2() API function. * {{{SUMS_USEMTSUMS_CONNECTION}}} - use the MT SUMS daemon for the SUM_open() and SUM_close() API functions. * {{{SUMS_USEMTSUMS_DELETESUS}}} - use the MT SUMS daemon for the SUM_delete_series() API function. * {{{SUMS_USEMTSUMS_GET}}} - use the MT SUMS daemon for the SUM_get() API function. * {{{SUMS_USEMTSUMS_INFO}}} - use the MT SUMS daemon for the SUM_infoArray() API function. * {{{SUMS_USEMTSUMS_PUT}}} - use the MT SUMS daemon for the SUM_put() API function. * {{{SUMSD_LISTENPORT}}} - the port that SUMS listens to for incoming requests. * {{{SUMSD_MAX_THREADS}}} - the maximum number of SUs that SUMS can process simultaneously. * {{{SUMSERVER}}} - the SUMS host machine; this is {{{<SUMS host>}}}. * {{{WEB_DBUSER}}} - the DRMS database user account that {{{cgi}}} programs access when they need to read from or write to database relations. The {{{__MAKE__}}} section: * {{{CFITSIO_INCS}}} - the path to the installed CFITSIO header files: this is {{{<CFITSIO install dir>/include}}}. * {{{CFITSIO_LIB}}} - the name of the CFITSIO library. * {{{CFITSIO_LIBS}}} - the path to the installed CFITSIO library files: this is {{{<CFITSIO install dir>/lib}}}. * {{{POSTGRES_INCS}}} - the path to the installed PostgreSQL header files: this is {{{<PostgreSQL install dir>/include}}}. * {{{POSTGRES_LIB}}} - the name of the PostgreSQL C API library (AKA {{{libpq}}}): this is always {{{pq}}}. * {{{POSTGRES_LIBS}}} - the path to the installed PostgreSQL library files: this is {{{<PostgreSQL install dir>/lib}}}. When installing NetDRMS updates, copy the existing {{{config.local}}} to the new {{{<NetDRMS install dir>}}} and edit the copy as needed, using the new {{{config.local.newstyle.template}}} to obtain information about parameters new to the newer release. Many of the parameter values have been determined during the previous steps of the installation process. Run the configuration script, {{{configure}}}, a csh shell script which is included in {{{<NetDRMS install dir>}}}. {{{ $ whoami <NetDRMS production user> $ cd <NetDRMS install dir> $ ./configure }}} {{{configure}}} reads {{{config.local}}} and uses the contained parameters to configure many of the NetDRMS features. It creates several directories in {{{<NetDRMS install dir>}}}: * {{{bin}}} - a directory that contains links to all executables in the DRMS code tree * {{{include}}} - a directory that contains links to all the header files in the DRMS code tree * {{{jsds}}} - a directory that contains links to all JSOC Series Definition (JSD) files * {{{lib}}} - a directory that contains links to all libraries in the DRMS code tree * {{{localization}}} - this directory contains project-specific make files and various files that contain the processed parameter information from {{{config.local}}} * {{{scripts}}} - a directory that contains links to all script files in the DRMS code tree * {{{_<architecture>}}} - a directory that contains all compiled files (object files, symbol files, binaries) in subdirectories that mirror the paths of the paths where the corresponding source code resides. For example, the source code for the {{{show_info}}} program resides in {{{<NetDRMS install dir>/base/util/apps/show_info.c}}}, and the {{{show_info}}} binary is placed in {{{<NetDRMS install dir>/_<architecture>/base/util/apps/show_info}}}. {{{<architecture>}}} was determined in a previous installation step. There are two possible names for {{{<architecture>}}}: {{{_linux_x86_64}}} and {{{_linux_avx}}} (for hosts that support Advanced Vector Extensions). Compile DRMS. To make the DRMS part of NetDRMS, run: {{{ $ whoami <NetDRMS production user> $ cd <NetDRMS install dir> $ make }}} As {{{<PostgreSQL superuser>}}}, create the DRMS database production user {{{<DRMS DB production user>}}}. Since PostgreSQL automatically attempts to use the Linux user name as the PostgreSQL user name when a connection attempt is made, use Linux user {{{<NetDRMS production user>}}} for database user {{{<DRMS DB production user>}}}. Create a password for this user, and also {{{GRANT}}} all database privileges to {{{<DRMS DB production user>}}} (this is necessary so that {{{<DRMS DB production user>}}} can create namespaces in the {{{netdrms}}} database): {{{ $ whoami <PostgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# CREATE ROLE <DRMS DB production user> WITH LOGIN; postgres=# ALTER ROLE <DRMS DB production user> WITH PASSWORD '<new password>'; postgres=# GRANT ALL ON DATABASE netdrms TO <DRMS DB production user>; postgres=# \q $ }}} As {{{<NetDRMS production user>}}}, create a {{{.pgpass}}} file. This file contains the PostgreSQL user account password, obviating the need to manually enter the database password each time a database connection attempt is made: {{{ $ whoami <NetDRMS production user> $ cd $HOME $ vi .pgpass i <PostgreSQL host>:*:*:<DRMS DB production user>:<new password> ESC :wq $ chmod 0600 .pgpass }}} As {{{<PostgreSQL superuser>}}}, run an {{{SQL}}} script, and a {{{perl}}} script (which executes several {{{SQL}}} scripts), both included in the NetDRMS installation, to create the {{{admin}}} and {{{drms}}} schemas and their relations, the {{{jsoc}}} and {{{sumsadmin}}} database users, data types, and functions: {{{ $ whoami <PostgreSQL superuser> # use psql to execute SQL script $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U postgres -f <NetDRMS install dir>/base/drms/scripts/NetDRMS.sql netdrms CREATE SCHEMA GRANT CREATE TABLE CREATE TABLE GRANT GRANT CREATE SCHEMA GRANT CREATE ROLE CREATE ROLE $ perl <NetDRMS install dir>/base/drms/scripts/createpgfuncs.pl netdrms }}} For more information about the purpose of these objects, read the comments in the {{{NetDRMS.sql}}} and {{{createpgfuncs.pl}}}. Although we'd like for the {{{<DRMS DB production user>}}} to own all these database objects that were just created by these scripts (these objects are all specific to NetDRMS), we had to run the scripts as {{{<PostgreSQL superuser>}}}, so the superuser owns them at this point. This was necessary since {{{<DRMS DB production user>}}} lacks the elevated privileges necessary to create these database objects. Run the following, as {{{<PostgreSQL superuser>}}} to alter ownerships: {{{ $ whoami <PostgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> netdrms << EOF ALTER SCHEMA admin OWNER TO <DRMS DB production user>; ALTER TABLE admin.ns OWNER TO <DRMS DB production user>; ALTER TABLE admin.sessionns OWNER TO <DRMS DB production user>; ALTER SCHEMA drms OWNER TO <DRMS DB production user>; ALTER TYPE drmskw OWNER TO <DRMS DB production user>; ALTER TYPE drmsseries OWNER TO <DRMS DB production user>; ALTER TYPE drmssession OWNER TO <DRMS DB production user>; ALTER TYPE drmssg OWNER TO <DRMS DB production user>; ALTER TYPE rep_item OWNER TO <DRMS DB production user>; EOF ALTER SCHEMA ALTER TABLE ALTER TABLE ALTER SCHEMA ALTER TYPE ALTER TYPE ALTER TYPE ALTER TYPE ALTER TYPE $ }}} We recommend using the NetDRMS database user {{{<DRMS DB production user>}}} as the SUMS database user {{{<SUMS DB production user>}}}. However, feel free to create a new user if necessary. If the DRMS and SUMS databases reside in different clusters, then you ''will'' need to create the {{{<SUMS DB production user>}}}. Again, since PostgreSQL automatically attempts to use the Linux user name as the PostgreSQL user name when a connection attempt is made, use Linux user {{{<NetDRMS production user>}}} for database user {{{<SUMS DB production user>}}}. If you choose a {{{<SUMS DB production user>}}} that is not {{{<NetDRMS production user>}}}, then you will need to pass {{{<SUMS DB production user>}}} to both Remote SUMS and SUMS when starting them. {{{ $ # DO THIS ONLY IF <SUMS DB production user> != <DRMS DB production user> OR IF <DRMS database cluster> != <SUMS database cluster> $ whoami <PosgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# CREATE ROLE <SUMS DB production user>; postgres=# \q $ }}} In addition, you will need to create a SUMS database user that has read-only access to the SUMS database objects: {{{ $ whoami <PosgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# CREATE ROLE <SUMS DB readonly user> WITH LOGIN; postgres=# ALTER ROLE <SUMS DB readonly user> WITH PASSWORD 'readonlyuser'; postgres=# \q }}} where {{{<SUMS DB readonly user>}}} is {{{[SUMS_READONLY_DB_USER]}}}. This database account is used by the Remote SUMS Client, a daemon used to manage the auto-download of SUs for subscriptions. Remote SUMS Client will be run by {{{<NetDRMS production user>}}}, so add the password to {{{<NetDRMS production user>}}}'s {{{.pgpass}}} file: {{{ $ whoami <NetDRMS production user> $ cd $HOME $ vi .pgpass i <PostgreSQL host>:*:*:[SUMS_READONLY_DB_USER]:readonlyuser ESC :wq $ chmod 0600 .pgpass }}} If you created a new SUMS DB production user, add a password for this user. ''Ensure'' that you use the same password that you used for {{{<DRMS DB production user>}}} - you will use the same Linux user when connecting to either database, so the same {{{.pgpass}}} file will be used for authentication. As {{{<PostgreSQL superuser>}}}, run {{{psql}}} to add a password for this new database user: {{{ $ # DO THIS ONLY IF <SUMS DB production user> != <DRMS DB production user> $ whoami <PostgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# ALTER ROLE <DRMS DB production user> WITH PASSWORD '<DRMS DB production user password>'; postgres=# \q $ }}} SUMS stores directory and file information in relations in the SUMS database. To create those relations and initialize tables, as {{{<NetDRMS production user>}}} run: {{{ $ whoami <NetDRMS production user> $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U <SUMS DB production user> -f <NetDRMS root>/base/sums/scripts/postgres/create_sums_tables.sql netdrms_sums CREATE TABLE CREATE INDEX CREATE INDEX GRANT CREATE TABLE GRANT CREATE TABLE GRANT CREATE INDEX CREATE INDEX CREATE INDEX CREATE INDEX CREATE TABLE GRANT CREATE TABLE CREATE INDEX GRANT CREATE SEQUENCE GRANT CREATE SEQUENCE GRANT CREATE TABLE GRANT CREATE TABLE GRANT CREATE TABLE GRANT $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U <SUMS DB production user> netdrms_sums netdrms_sums=> ALTER SEQUENCE sum_ds_index_seq START <min val> RESTART <min val> MINVALUE <min val> MAXVALUE <max val>; ALTER SEQUENCE netdrms_sums=> \q $ }}} where {{{<min val>}}} is {{{<drms site code> << 48}}}, and {{{<max val>}}} is {{{<min val> + <maximum unsigned 48-bit integer> - 1}}}, where {{{<drms site code>}}} is the value of the {{{[DRMS_LOCAL_SITE_CODE]}}}, and <maximum unsigned 48-bit integer> is 2^48^ (which is {{{281474976710656}}}). For the JSOC (site code 0x0000), this {{{ALTER SEQUENCE}}} command looks like: {{{ netdrms_sums=> ALTER SEQUENCE sum_ds_index_seq START 0 RESTART 0 MINVALUE 0 MAXVALUE 281474976710655; }}} You now have the base part of NetDRMS installed. This excludes projects that are private to other NetDRMSs, such as the JSOC. It is often the case that your team would like to use private code as part of a data-generation pipeline. For example, the JSOC maintains time-distance analysis code that is part of the JSOC DRMS code tree, but not part of the base NetDRMS installation. To obtain this code, you can install those projects into your existing NetDRMS by: 1. obtaining the projects from the its source (like the JSOC), and placing it in your {{{<NetDRMS install dir>/proj}}} directory; and 1. modifying your {{{config.local}}} to refer to new projects, running {{{configure}}}, and then running {{{make}}} <<Anchor(initialize-sums-disk)>> == Initializing SUMS Storage == In addition to SUMS database relations, SUMS requires a file system on which SUMS maintains storage areas called ''SUMS partitions''. A SUMS paritition is really just a directory that contains SUMS Storage Units (each of which is implemented as a subdirectory inside the SUMS partition). === New Storage === As {{{<NetDRMS production user>}}} create one or more partitions now - although we have had success making them as large as 60 TB, make 40 TB partitions. For example, if you plan on setting aside X TB of SUMS storage, then make approximately {{{N = <total storage TB> / X}}} 40 TB partitions. The partitions can reside on a file server and be mounted onto all machines that will use NetDRMS, but the following example simply creates directories on a single file system on {{{<SUMS partition host>}}}. First, make the root directory that contains the SUMS paritions (something like {{{/opt/sums}}}): {{{ $ sudo mkdir <SUMS root> # allow SUMS users to write into SUMS $ sudo chown <NetDRMS production user>:<SUMS users> <SUMS root> # when a user writes a file into SUMS, make sure that the file's group owner is <SUMS users> $ sudo chmod g+s <SUMS root> }}} {{{<SUMS users>}}} is the Linux group that is allowed to write to SUMS. You created it in a previous step. Second, make the SUMS partitions: {{{ $ whoami <NetDRMS production user> $ hostname <SUMS partition host> $ mkdir -p <SUMS root>/partition01 $ mkdir -p <SUMS root>/partition02 ... $ mkdir -p <SUMS root>/partitionN }}} Initialize the SUMS DB {{{sum_partn_avail}}} table with the names of these partitions. For each SUMS partition run the following: {{{ $ whoami <NetDRMS production user> $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U <SUMS DB production user> netdrms_sums netdrms_sums=> INSERT INTO sum_partn_avail (partn_name, total_bytes, avail_bytes, pds_set_num, pds_set_prime) VALUES ('<SUMS partition path>', <avail bytes>, <avail bytes>, 0, 0); }}} where {{{<SUMS partition path>}}} is the full path of the partition as seen from {{{<SUMS host>}}} (which is where the SUMS daemon will run) and {{{<avail bytes>}}} is some number less than the number of bytes in the directory (multiply the number of blocks in the directory by the number of bytes per block). The number does not matter, as long as it is not bigger than the total number of bytes available. SUMS will adjust this number as needed. === Existing Storage === It might be the case that you would like to have a new NetDRMS installation use existing SUMS partitions. To do this, you would enter the paths to the existing partitions into the {{{sum_partn_avail}}} table. At this point, the SUMS database tables that contain the ''pointers'' to SUs ({{{sum_main}}} and {{{sum_partn_alloc}}}) do not contain any references to the existing SUs. Most likely the existing SUMS partitions are part of an existing, old NetDRMS, in which case these references are in that old NetDRMS SUMS database table. If so, then you can manually ''dump'' four database tables/sequences from the old NetDRMS to files that can be ingested into the new NetDRMS installation. To do that, as the old NetDRMS {{{<NetDRMS production user>}}} run the following: {{{ $ whoami <NetDRMS production user> $ psql -h <OLD NetDRMS host> -p <OLD NetDRMS port> netdrms_sums netdrms_sums=> COPY public.sum_main TO '/tmp/sum_main_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_partn_alloc TO '/tmp/sum_partn_alloc_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_seq TO '/tmp/sum_seq_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_ds_index_seq TO '/tmp/sum_ds_index_seq_dump.txt' WITH ENCODING 'UTF8'; }}} The {{{COPY}}} command will save the table data onto a file on the OLD NetDRMS PostgreSQL server. Copy those files to the NEW NetDRMS {{{<PostgreSQL host>}}} and then ingest them: {{{ $ whoami <NetDRMS production user> $ psql -h <NEW NetDRMS host> -p <NEW NetDRMS port> netdrms_sums netdrms_sums=> COPY public.sum_main FROM '/tmp/sum_main_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_partn_alloc FROM '/tmp/sum_partn_alloc_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_seq FROM '/tmp/sum_seq_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_ds_index_seq FROM '/tmp/sum_ds_index_seq_dump.txt' WITH ENCODING 'UTF8'; }}} <<Anchor(create-users)>> == Creating DRMS User Accounts == For each Linux user, {{{<new DRMS user>}}} who will run installed DRMS modules, or who will write new DRMS modules, you will need to set up their environment, create their ''DRMS account'', and create their {{{.pgpass}}} file so they can run DRMS modules without having to manually authenticate to the database. Just like you did for the {{{<NetDRMS production user>}}}, you will need to unset the {{{PYTHONPATH}}} environment variable, and put PostgreSQL executables into the user's {{{$PATH}}}. You will also need to put the {{{<NetDRMS production user>}}}'s {{{netdrms}}} Python virtual environment into the user's {{{$PATH}}} - each user will use this Python environment, not the base installation. Also set the {{{JSOCROOT}}}, {{{JSOC_MACHINE}}}, {{{COMPILER}}}, and {{{FCOMPILER}}} environment variables: {{{ # .bashrc # PostgreSQL executables export PATH=/usr/pgsql-12/bin:$PATH # python production virtual environment unset PYTHONPATH export PATH=<NetDRMS production home>/.conda/envs/netdrms/bin:$PATH # NetDRMS binary path export PATH=<NetDRMS root>/bin/<architecture>:$PATH # root of NetDRMS source/installation export JSOCROOT=<NetDRMS root> # architecture export JSOC_MACHINE=<architecture> # set COMPILER to icc for the Intel C++ compiler, and to gcc for the GNU C++ compiler export COMPILER=<C compiler> # set to ifort for the Intel Fortran compiler, and to gfortran for the GNU Fortran compiler export FCOMPILER=<Fortan compiler> }}} You will also need to add them to the {{{<SUMS users>}}} group: {{{ $ sudo usermod -a -G <SUMS users> <new DRMS user> }}} To create a DRMS account, you create a database account for the user, plus you add user-specific rows to various DRMS database tables. The script {{{newdrmsuser.pl}}} exists to facilitate these tasks: {{{ $ whoami <NetDRMS production user> $ perl <NetDRMS root>/base/drms/scripts/newdrmsuser.pl netdrms <PostgreSQL host> <PostgreSQL port> <new DRMS user> <initial password> <new DB user namespace> user 1 Connection to database with 'dbi:Pg:dbname=netdrms;host=drms;port=<PostgreSQL port>' as user '<new DRMS user>' ... success! executing db statment ==> CREATE USER <new DRMS user> executing db statment ==> ALTER USER <new DRMS user> WITH password '<initial password>' executing db statment ==> GRANT jsoc to <new DRMS user> running cmd-line ==> masterlists dbuser=<new DRMS user> namespace=<new DB user namespace> nsgrp=user Please type the password for database user "postgres": Connected to database 'netdrms' on host '<PostgreSQL host>' and port '<PostgreSQL port>' as user 'postgres'. Created new drms_series... Created new 'drms_keyword'... Created new 'drms_link'... Created new 'drms_segment'... Created new 'drms_session'... Created new drms_sessionid_seq sequence... Commiting... Done. executing db statment ==> INSERT INTO admin.sessionns VALUES ('<new DRMS user>', '<new DB user namespace>') }}} where {{{<new DB user namespace>}}} is the PostgreSQL namespace dedicated to the new user. A namespace is a logical container that allows a database user to own database objects, like relations, that have the same name as objects owned by other users - items in a namespace need only be uniquely named within the namespace, not between namespaces. For example, the relation {{{drms_series}}} in the namespace {{{su_arta}}} is not the same relation as the {{{drms_series}}} relation in the {{{su_phil}}} namespace - the relations have the same name, but they are different relations. In virutally all PostgreSQL operations, a user can prefix the name of a relation with the namespace: {{{su_arta.drms_series}}} refers to the first relation, and {{{su_phil.drms_series}}} refers to the second relation. The purpose of {{{<new DB user namespace>}}} is to hold non-production, private data series - sort of a private user space to develop new DRMS modules to create data. If those data should become a production-level products, then the data and the code that generates the data need to be moved to a production namespace. At the JSOC, we have several such production namespaces (e.g., {{{aia}}}, {{{hmi}}}, {{{mdi}}}). A site creates production namespaces with a different module ({{{masterlists}}}; {{{newdrmsuser.pl}}} is only for creating non-production namespaces. Please see the NOTE in [[http://jsoc.stanford.edu/jsocwiki/NewDrmsUser|this page]] for assistance with choosing {{{<new DB user namespace>}}}. The general naming convention is to prepend the namespace with an abbreviation to identify the site that owns the data in the namespace. For example, all private data created at Stanford reside in dataseries whose namespaces start with {{{su_}}} (Stanford University), regardless of the affiliation of the user who creates data in this namespace. Data created at NASA Ames start with {{{nas_}}} (NASA Supercomputing Division). Following the underscore is a string to identify a particular user - {{{su_arta}}} for Art, and {{{su_phil}}} for Phil. You can also specify a group with the suffix (e.g., {{{su_uscsolar}}} for a solar group at the University of Southern California that creates data at Stanford. {{{<initial password>}}} is the initial password for this account - the initial password does not matter much since you are going to have the user change it next. Running {{{newdrmsuser.pl}}} will create a new DRMS database user that has the same name as the user's Linux account name. Have the user change their password: {{{ $ whoami <new DRMS user> $ psql -h <PostgreSQL host> -p <PostgreSQL port> netdrms netdrms=> ALTER USER <new DRMS user> WITH PASSWORD '<new password>'; netdrms=> \q $ }}} And then have the user create their {{{.pgpass}}} file (to allow auto-login to their database account) and set permissions to {{{0600}}}: {{{ $ whoami <new DRMS user> $ cd $HOME $ vi .pgpass i <PostgreSQL host>:*:*:<DRMS DB production user>:<new password> ESC :wq $ chmod 0600 .pgpass }}} Please click [[http://jsoc.stanford.edu/jsocwiki/DrmsPassword|here]] for additional information on the {{{.pgpass}}} file. If you plan on creating data that will be publicly distributed, you should also create one or more data-production users. For example, if you plan on making a new public HMI data series, you could create a user named {{{hmi_production}}}. Although you could follow previous steps to create a new Linux account for this database user, you do not necessarily need to. Instead you can use the existing <NetDRMS production user> and have it connect as the {{{hmi_production}}} user. To do that, first create the new {{{hmi_production}}} database user by running {{{newdrmsuser.pl}}} as just described. Choose a descriptive namespace that follows the naming guidelines described about, like {{{hmi}}}. Because a {{{.pgpass}}} already exists for <NetDRMS production user>, you want to '''''ADD''''' a new line to {{{.pgpass}}} for this user. Continuing with the {{{hmi_production}}} user example, add a password line for {{{hmi_production}}}: {{{ # .pgpass <PostgreSQL host>:*:*:hmi_production:<hmi_production password> }}} At this point, it is a good idea to test that the new DRMS user can use a basic NetDRMS program. Although your NetDRMS has no DRMS data series, running {{{show_series}}} is a good way to test various components, like authentication, database connection, etc. Test DRMS by running the show_series command: {{{ $ whoami <new DRMS user> $ show_series $ }}} Nothing will be printed when this command is run, since your NetDRMS is devoid of data at the moment, but if you see no errors, then life is good. If not, then contact the JSOC for help troubleshooting. To perform a more thorough test involving SUMS, you will need to have a least one DRMS data series that has SUMS data. You can obtain such a data series by [[#register-subscriptions|registering for subscriptions]]. <<Anchor(run-sums)>> == Running SUMS Services == Before you can use NetDRMS, you, as {{{<NetDRMS production user>}}} on {{{<SUMS host>}}}, will need to start SUMS. To launch one or more SUMS daemons, {{{sumsd.py}}}, use the {{{start-mt-sums.py}}} script: {{{ $ whoami <NetDRMS production user> $ hostname <SUMS host> $ python3 <NetDRMS root>/base/sums/scripts/start-mt-sums.py daemon=<NetDRMS root>/base/sums/scripts/sumsd.py ports=6102 --loglevel=debug running /home/netdrms_production/.conda/envs/netdrms/bin/python3 /opt/netdrms/base/sums/scripts/sumsd.py --sockport=6102 --loglevel=debug started instance /opt/netdrms/base/sums/scripts/sumsd.py:6102 (pid 27515) {"started": [27515]}}}} This command starts {{{sumsd.py}}}, and creates the ''instances'' file and a log file in {{{SUMLOG_BASEDIR}}} by default. The instances file is a state file used by {{{start-mt-sums.py}}} and {{{stop-mt-sums.py}}} to manage the running {{{sumsd.py}}} instances. By default, the log file is named {{{sumsd-<PPPPP>-<YYYYMMDD.HHMMSS>.txt}}}, where {{{<PPPPP>}}} is the PID of the {{{sumsd.py}}} process, and {{{<YYYYMMDD.HHMMSS>}}} is the time string representing the time the instance was launched. The complete usage is: {{{ usage: start-mt-sums.py daemon=<path to daemon> ports=<listening ports> [ --instancesfile=<instances file path> ] [ --loglevel=<critical, error, warning, info, or debug>] [ --logfile=<file name> ] [ --quiet ] optional arguments: -h, --help show this help message and exit -i <instances file path>, --instancesfile <instances file path> the json file which contains a list of all the sumsd.py instances running -l LOGLEVEL, --loglevel LOGLEVEL specifies the amount of logging to perform; in order of increasing verbosity: critical, error, warning, info, debug -L <file name>, --logfile <file name> the file to which sumsd logging is written -q, --quiet do not print any run information required arguments: d <path to daemon>, daemon <path to daemon> path of the sumsd.py daemon to launch p <listening ports>, ports <listening ports> a comma-separated list of listening-port numbers, one for each instance to be spawned }}} {{{start-mt-sums.py}}} will fork one or more {{{sumsd.py}}} daemon processes. The {{{ports}}} argument identifies the SUMS host ports on which sumsd.py will listen for client (DRMS module) requests. One sumsd.py process will be invoked per port specified. Each process creates a log-file in {{{[SUMLOG_BASEDIR]}}} named, by default, {{{sumsd-<port>-<YYYYMMDD>.<HHMMSS>.txt}}}. The {{{-L/--logfile}}} argument allows you to override the path and name to this log file. To stop one or more SUMS services, use the {{{stop-mt-sums.py}}} script: {{{ $ whoami <NetDRMS production user> $ hostname <SUMS host> $ python3 <NetDRMS root>/base/sums/scripts/stop-mt-sums.py daemon=<NetDRMS root>/base/sums/scripts/sumsd.py --ports=6102 }}} The complete usage is: {{{ usage: stop-mt-sums.py [ -h ] daemon=<path to daemon> [ ---ports=<listening ports> ] [ --instancesfile=<instances file path> ] [ --quiet ] optional arguments: -h, --help show this help message and exit -p <listening ports>, --ports <listening ports> a comma-separated list of listening-port numbers, one for each instance to be stopped -i <instances file path>, --instancesfile <instances file path> the json file which contains a list of all the sumsd.py instances running -q, --quiet do not print any run information required arguments: d <path to daemon>, daemon <path to daemon> path of the sumsd.py daemon to halt }}} <<Anchor(register-subscriptions)>> == Registering for Subscriptions == A NetDRMS site can optionally register for a data-series subscription to any NetDRMS site that offers subscription service. The JSOC NetDRMS offers subscriptions, but at the time of this writing, no other site does. Once a site registers for a data series subscription, the site will become a mirror for that data series. The subscription process ensures that the mirroring site will receive regular and timely updates made to the data series by the serving site. The subscribing site can configure the interval between updates such that the mirror can synchronize with the server and receive updates within a couple of minutes, keeping the mirror up-to-date in (almost) real time. To register for a subscription, {{{<Subscription production user>}}} will set up {{{ssh}}} keys (for SU transfer), start daemons, and run the subscription script, {{{subscribe.py}}}. The assumption is that {{{<Subscription production user>}}} is {{{<NetDRMS production user>}}}, but you are free to choose a different user. {{{subscribe.py}}} makes subscription requests to the serving site's subscription manager. The process entails the creation of a snapshot of the data-series DRMS database information at the serving site. Those data are downloaded, via HTML, to the subscribing site, where they are ingested by {{{subscribe.py}}}. {{{get_slony_logs.pl}}}, a client-side cron task, updates the data-series snapshot with any server-side changes that have been made since the snapshot was created. Ingestion of the snapshot results in the creation of the DRMS database objects that maintain and store the data series, and {{{get_slony_logs.pl}}} updates those objects when the server makes changes. At this time, no SUMS data files are downloaded. Instead, and optionally, the IDs for the series' SUMS Storage Units (SU) are saved in a database relation. It is the function of Remote SUMS ({{{rsums.py}}}), another NetDRMS daemon, to download the SUs and ingest them into your SUMS. Remote SUMS accepts ''requests'' from DRMS for SUs. It then communicates with the serving NetDRMS and manages the {{{scp}}} download and ingestion of those SUs. Once Remote SUMS is running, should any DRMS code/module request an SU that is not present in the local NetDRMS, DRMS will send a download request to Remote SUMS. The Remote SUMS Client ({{{rsums-clientd.py}}}), an optional NetDRMS daemon, can automate this process so that when new subscription data are ingested into the DRMS database, it submits requests for the associated SUs to Remote SUMS. In this way, it is possible to ''pre-fetch'' the SU files before any user requests them. But pre-fetching is optional. The SUs will be downloaded on-demand as described above. In fact, if the subscribing NetDRMS site were to automatically download an SU, then delete the SU (there is a method to do this, described later), then an on-demand download is the only way to re-fetch the deleted SU. On-demand downloads happen automatically; any DRMS module that attempts to access an SU (like with a {{{show_info -p}}} command) that is not present for any reason will trigger an {{{rsumsd.py}}} request. The module will pause until the SU has been downloaded, then automatically resume its operation on the previously missing SU. As {{{get_slony_logs.pl}}} users {{{scp}}} to download update files (''Slony logs'') and {{{rsumsd.py}}} uses {{{scp}}} to automatically download SUs, SSH public-private keys must be created at the subscribing site, and the public key must be provided to the serving site. Setting this up requires coordinated work at both the susbscribing and serving sites. As {{{<Subscription production user>}}} on the subscribing site, run {{{ $ whoami <Subscription production user> $ ssh-keygen -t rsa }}} This will allow you to create a passphrase for the key. If you choose to do this, then save this phrase for later steps. In the home directory of {{{<Subscription production user>}}}, {{{ssh-keygen}}} will create a public key named {{{id_rsa.pub}}}. Provide this public key to the serving site. The serving site must then add the public key to its list of authorized keys. If the {{{.ssh}}} directory does not exist, then the serving site must first create this directory and give it {{{0700}}} permissions. If the {{{authorized_keys}}} file in {{{.ssh}}} does not exist, then it must first be created and given {{{0644}}} permissions: {{{ $ whoami <Subscription manager user> $ mkdir .ssh $ chmod 0700 .ssh $ cd .ssh $ touch authorized_keys $ chmod 0644 authorized_keys }}} Once the {{{.ssh}}} and {{{authorized_keys}}} files exist and have the proper permissions, the serving site administrator can then add the client site's public key to its list of authorized keys: {{{ $ whoami <Subscription manager user> $ cd $HOME/.ssh $ cat <remote-site public key file> >> authorized_keys }}} Back at the client NetDRMS site, if an {{{ssh}}} passphrase was chosen, then as {{{<Subscription production user>}}} start an {{{ssh-agent}}} instance to automate the passphrase authentication. If no passphrase was provided when {{{ssh-keygen}}} was run, this step can be skipped. Otherwise, run: {{{ $ whoami <Subscription production user> $ ssh-agent > $HOME/.ssh-agent $ source $HOME/.ssh-agent # needed for ssh-add, and also for rsumsd.py and get_slony_logs.pl $ ssh-add $HOME/.ssh/id_rsa }}} and provide the passphrase. To keep ingested data series synchronized with changes made to it at the serving site, a client-side cron tab runs {{{get_slony_logs.pl}}} periodically. This perl script uses {{{scp}}} to download ''slony log files'' - SQL files that insert, delete, or update database relation rows. {{{get_slony_logs.pl}}} communicates with the Slony-I replication software running at the serving site. Slony-I generates these log (SQL) files at the server which are then downloaded by the client. To register for a subscription to a new series, run: {{{ $ python3 <NetDRMS root>/base/drms/replication/subscribe_series/subscribe.py cfg=<subscription config file> reqtype=subscribe series=<published DRMS data series> --loglevel=debug }}} The complete usage is: {{{ usage: subscribe.py [ -hjpl ] cfg=<client configuration file> reqtype=<subscribe, resubscribe, unsubscribe> series=<comma-separated list of series> [ --archive=<0, 1>] [ --retention=<number of days>] [ --tapegroup=<tape-group number> ] [ --pg_user=<subscripton client DB user> ] [ --logfile=<log-file name> ] optional arguments: -h, --help show this help message and exit archive <series archive flag>, --archive <series archive flag> The tape archive flag for the series - either 0 (do not archive) or 1 (archive). retention <series SU disk retention>, --retention <series SU disk retention> The number of days the series SUs remain on disk before becoming subject to deletion. tapegroup <series SU tape group>, --tapegroup <series SU tape group> If the archive flag is 1, the number identifying the group of series that share tape files. pg_user <series SU tape group>, --pg_user <series SU tape group> The DB account the subscription client uses. -p, --pause Pause and ask for user confirmation before applying the downloaded SQL dump file. --loglevel LOGLEVEL Specifies the amount of logging to perform. In increasing order: critical, error, warning, info, debug --logfile <file name> The file to which logging is written. --filtersus <SU filter> Specifies a series keyword K and a number of days D, comma separated; a remote-SUMS request for an SU will occur only if the keyword K of the record containing the SU has a value that lies within the time interval determined by the days D. required arguments: cfg <client configuration file>, --config <client configuration file> The client-side configuration file used by the subscription service. reqtype <request type>, --reqtype <request type> The type of request (subscribe, resubscribe, or unsubscribe). series SERIES, --series SERIES A comma-separated list of DRMS series to subscribe/resubscribe to, or to unsubscribe from. }}} The debug level logging is not necessary, but if this is your first subscription, we recommend running with lots of debug statement - if there are issues, you can send this output to the JSOC for help. {{{<subscription config file>}}} contains parameters used by both the subscription client {{{subscribe.py}}} and the program that downloads Slony logs file, {{{get_slony_logs.pl}}}. We recommend making a directory {{{subscription}}} in the home directory of {{{<NetDRMS production user>}}} and saving the configuration file in that directory. The parameters to be set in {{{<subscription config file>}}} are as follows: * {{{node}}} - the name of the subscription client (e.g., {{{jsoc}}}, {{{nsocu}}}, {{{sdac}}}; must be globally unique across all NetDRMS sites; this string will be used in various state files and in file/directory names. * {{{kRSServer}}} - the full domain name of the subscription log server (e.g., {{{jsocport.stanford.edu}}} for a client subscibing to data series published by the JSOC). * {{{kRSUser}}} - the account on {{{kRSServer}}} that will be used for data transfer (e.g., {{{jsocexp}}} for a client subscibing to data series published by the JSOC). * {{{kRSPort}}} - the port on {{{kRSServer}}} that will be used for data transfer (e.g., {{{22}}} for {{{scp}}}). * {{{kRSBaseURL}}} - the base URL for all subscription services provided by subscription server for the Slony cluster (identified by {{{slony_cluster}}}) * {{{pg_host}}} - the client machine that hosts the client PostgreSQL database that will contain the replicated data series - this is {{{<PostgreSQL host>}}}. * {{{pg_port}}} - the port on the {{{pg_host}}} machine that will be used for communication with the data-series database - this is {{{<PostgreSQL port>}}}. * {{{pg_user}}} - the PostgreSQL user that will own the replicated series - this is {{{<NetDRMS production user>}}}. * {{{pg_dbname}}} - the name of the PostgreSQL database that resides on {{{pg_host}}} - this is {{{netdrms}}}. * {{{slony_cluster}}} - the name of the Slony cluster to which this node belongs (e.g., {{{jsoc}}} for a client subscribing to data series published by the JSOC). * {{{kLocalLogDir}}} - the client directory that will contain the subscription-process logs; we recommend {{{<NetDRMS production user home>/subscription/log}}}. * {{{kLocalWorkingDir}}} - the path to a directory for temporary working subscription files * {{{kSQLIngestionProgram}}} - the path to the script/program that will ingest the site-specific slony logs — usually the path to {{{get_slony_logs.pl}}} ({{{<NetDRMS root>/base/drms/replication/get_slony_logs.pl}}} * {{{kSubService}}} - the URL of the application at the subscription-serving site that accepts new subscription requests (for the JSOC subscripton server this is {{{${kRSBaseURL}/request-subs.py}}}) * {{{kPubListService}}} - the URL of the application at the subscription-serving site that that lists published data series (for the JSOC subscripton server this is {{{${kRSBaseURL}/publist.py}}}) * {{{kSubXfer}}} - the URL of the application at the subscription-serving site where subscription dump files are located (for the JSOC subscripton server this is {{{http://jsoc.stanford.edu/subscription}}}) * {{{kDeleteSeriesProgram}}} - the path to the program {{{delete_series}}}, which is used to delete DRMS data series on the client when requested ({{{<NetDRMS root>/bin/<architecture>/delete_series}}}). * {{{archive}}} - for new subscriptions, the default data series archive action; set to 1 if the NetDRMS site has a tape archive system AND the default is to archive all series obtained by subscription * {{{retention}}} - for new subscriptions, the default number of days to retain SUs (after this many days SUs are marked for deletion and subjected to garbage collection as needed) * {{{tapegroup}}} - for new subscriptions, the default archive tape group for the the series' SUs (ignored if archive == 0) * {{{ingestion_path}}} - the local directory that will contain the ingestion "die" file — used by {{{get_slony_logs.pl}}} * {{{scp_cmd}}} - the absolute path to the client's scp program. * {{{ssh_cmd}}} - the absolute path to the client's ssh program. * {{{rmt_slony_dir}}} - the absolute path, accessible from the {{{kRSUser}}} account, on the server to the directory that contains the site-specific slony logs (/data/pgsql/slon_logs/live/site_logs). * {{{slony_logs}}} - the client directory that contains the downloaded site-specific slony logs. * {{{PSQL}}} - the path to the client's {{{psql}}} program, and any flags needed to run psql as the pg_user user, like -h {{{pg_host}}}. * {{{email_list}}} - the email account to which error messages will be sent. You may find that a subscription has gotten out of sync, for various reasons, with the serving site's data series (accidental deletion of database rows, for example). You can re-register for the subscription to true-up. The existing DRMS data will be deleted, and replaced with a fresh snapshot. You will be prompted asking if you would like to delete the series' SUMS data (SUs, FITS files, etc.). If you are sure you no longer need them, go ahead and say ''yes''. {{{ python <NetDRMS root>/base/drms/replication/subscribe_series/subscribe.py cfg=<subscription config file> reqtype=resubscribe series=<subscription series> --loglevel=debug }}} Finally, there might come a time where you no longer which to hold on to a registration. To remove the subscription from your set of registered data series run: {{{subscribe.py}}} can be used to alleviate this problem. Run the following to re-do the subscription registration: {{{ python <NetDRMS root>/base/drms/replication/subscribe_series/subscribe.py cfg=<subscription config file> reqtype=unsubscribe series=<subscription series> --loglevel=debug }}} NOTE: In this case, the {{{subscribe.py}}} will prompt you to ask if you would like to delete the existing DRMS data series. The answer is usually ''yes''. However, if you need to keep the existing data series snapshot for some reason (e.g., you want to work with the existing data, but simply do not want to ingest new data), then respond ''no''. Keep in mind that you will not be able to register for a subscription to the same series if that series already exists - you will need to run {{{delete_series}}}. In all the above commands, the logging level is set to {{{debug}}}. It is a good idea to enable verbose logging like this the first time you run one of these commands, just in case an issue occurs (usually do to some configuration issue). Providing the debug log to the JSOC when troubleshooting will be invaluable. Once you have successfully registered for subscription to at least one series, you will need to install a crontab to run {{{get_slony_logs.pl}}}, which updates the subscription series, on a regular basis. {{{get_slony_logs.pl}}} has a dependency on the {{{Net::SSH}}} perl package. If you are running CentOS, then there is a package available to install this package: {{{ }}} If you cannot find a package, then you can use [[ https://metacpan.org/pod/Net::SSH | CPAN ]]. Once you have the {{{Net::SSH}}} Perl package installed, run {{{subscribe.py}}} as {{{<Subscription production user>}}}: {{{ $ whoami <Subscription production user> $ crontab -e */5 * * * * (. ~/.ssh-agent; <NetDRMS root>/base/drms/replication/get_slony_logs.pl <subscription config file> >> ~/<path>/get_slony_logs.log 2>&1 ) }}} At the JSOC, Slony log files are created every minute. It is not the case that your series' subscriptions will have been updated every minute (this depends on the cadence of the series to which you have subscriptions), but checking every minute will minimize the lag between the state of your series and the state of your series at the serving site. <<Anchor(run-remote-sums)>> == Running Remote SUMS == Your NetDRMS may contain data produced by other, non-local NetDRMSs. Via a variety of means, the local NetDRMS can obtain and ingest the database information for these data series produced non-locally. In order to use the associated data files (typically image files), the local NetDRMS must download the storage units (SUs) associated with these data series too. Remote SUMS, a tool that comes with NetDRMS, downloads SUs as needed - i.e., if a DRMS module or program requests the path to the SU or attempts to read it, and it is not present in the local SUMS yet, Remote SUMS will download the SUs. While the SUs are being downloaded, the initiating module or program will poll waiting for the download to complete. Several components compose Remote SUMS. On the client side, the local NetDRMS, is a daemon that must be running {{{rsumsd.py}}}. There also must exist some database tables, as well as some binaries used by the daemon. On the server side, all NetDRMS sites that wish to act as a source of SUs for the client, is a CGI (rs.sh). This CGI returns file-server information (hostname, port, user, SU paths, etc.) for the SUs the server has available in response to requests that contain a list of SUNUMs. When the client encounters requests for remote SUs that are not contained in the local SUMS, it sends a request the Remote SUMS daemon to download those SUs. The client code then polls waiting for the request to be serviced. The daemon in turn sends requests to all rs.sh CGIs at all the relevant providing sites. The owning sites return the file-server information to the daemon, and then the daemon downloads the SUs the client has requested, via scp, and notifies the client module once the SUs are available for use. The client module will then exit from its polling code and continue to use the freshly downloaded SUs. To use Remote SUMS, several {{{config.local}}} parameters must be present. If you followed the steps in this document, then you have already set those parameters. Please see [[#install-netdrms | Installing NetDRMS]] for a description of each one. Each SU that is downloaded has an associated expiration date, a flag indicating whether or not the SU is archived, and if the SU is archived the tapegroup to which the SU belongs. The manner in which the values for these parameters are determined is a bit complicated. When you register for a series subscription, the series is created at your site, and at that time, the values for these parameters are determined. The series is then initialized with these values. There are default value for these parameters that are overridden by optional parameters supplied during the call to {{{subscribe.py}}}. In the following list, the lower-numbered items, if present, override the higher-numbered items, in this order: 1. the {{{--archive, --retention, --tapegroup}}} command-line arguments to {{{subscribe.py}}} 1. the {{{archive, retention, tapegroup}}} parameters in {{{subscription config file}}} Now, when Remote SUMS runs, the values of the parameters for SUs downloaded and ingested are determined in a similar hierarchical fashion, with higher-numbered items overriding lower-numbered items: 1. the parameters associated with the series, as determined above 1. the {{{--archive, --expiration, --lifespan, --tapegroup}}} command-line arguments to {{{rsumsd.py}}} 1. {{{[RS_SU_ARCHIVE], [RS_SU_EXPIRATION], [RS_SU_LIFESPAN], [RS_SU_TAPEGROUP]}}} To run Remote SUMS, as {{{NetDRMS production user}}} run the following to create the requests database table: {{{ $ whoami <NetDRMS production user> $ python3 <NetDRMS root>/base/drms/scripts/rscreatetabs.py tabs=req }}} Remote SUMS downloads SUs via {{{scp}}}. As such, you will need to create SSH keys, distribute the public one to the site serving the SUs, and start up an {{{ssh-agent}}} if you have not already done so - you should have already though since it is needed by {{{get_slony_logs.pl}}}, a component of the subscription system (please see [ [[#register-subscriptions|Registering for Subscriptions]] ]). To launch Remote SUMS, ensure {{{sumsd.py}}} is first running, then as {{{<NetDRMS production user>}}} {{{source}}} the {{{ssh-agent}}} environment file, and then run: {{{ $ whoami <NetDRMS production user> $ source $HOME/.ssh-agent $ python <NetDRMS root>/base/drms/scripts/rsumsd.py --loglevel=debug }}} For now, we recommend setting the log level to debug; when things appear to be running smoothly, then you can restart with the default level (info). The output log named {{{<rslog_YYYYMMDD.txt>}}} will be written to the directory identified by {{{[RS_LOGDIR]}}}, so make sure that directory exists before running {{{rsumsd.py}}}. <<Anchor(#run-remote-sums-client)>> == Running Remote SUMS Client == If you have at least one subscription, and you have Remote SUMS running (''not'' the JMD), then you can automate the download of SUs for the subscriptions. This is optional, however. If you skip this step, then when a NetDRMS user attempts to use one or more SUs that are part of a subscription (and hence not at your NetDRMS site initially), the SUs will be downloaded in an on-demand fashion. However, it may be desirable to pre-fetch the SUs if certain usage patterns hold. If one or more users are going to use a block of SUs for a two-week period, for example, then downloading hundreds or thousands of SUs one at a time would be very inefficient. Also, it is very common for NetDRMS users to use newly produced SUs soon after they are created at the subscription server. In this case, a good strategy might be to automatically download the latest SUs for all subscriptions, knowing that lots of very recent SUs will be popular. The Remote SUMS Client, {{{rsums-clientd.py}}}, monitors SU references in incoming Slony logs. It groups SUs into batches, and submits requests containing these batches to Remote SUMS ,{{{rsumsd.py}}}. {{{rsums-clientd.py}}} then monitors the progress of these requests, logging the results. In addition to a running {{{rsumsd.py}}}, {{{rsums-clientd.py}}} requires the existence of three components. The ''capture table'' is a database table that contains a list of all SUNUMs that are to be downloaded. It gets automatically populated as Slony logs are ingested. The ''capture function'' is a database function that performs the actual work of inserting rows into the capture table. It is called every time a Slony log inserts a row into a subscribed series' series table. You will need to create a ''capture trigger'' database trigger for each series table for which you want to automate SU downloads. These components work together as follows: * each capture trigger watches a series table, and when a row is inserted (due to {{{get_slony_logs.pl}}} ingesting new Slony logs), the trigger runs the capture function * this function then inserts a row into the capture table * {{{rsums-clientd.py}}} then ''sees'' the newly inserted capture-table row and extracts the SUNUM * {{{rsums-clientd.py}}} batches several of these SUNUMs, and then makes a {{{rsumsd.py}}} request out of them * {{{rsumsd.py}}} processes the requests (multiple in parallel), downloading one or more SUs and ingesting them into the local SUMS * {{{rsumsd.py}}} updates a status column in the capture table * {{{rsums-clientd.py}}} reads the status, and upon success, it removes the rows that contained the SUs that were sucessfully ingested; upon failure, {{{rsums-clientd.py}}} will log an error, and can optionally re-try one or more times To use {{{rsums-clientd.py}}}, {{{<NetDRMS production user>}}} must first create the capture table and the capture function. They must also add the capture trigger to the series table of the DRMS data series for which they want to enable automatic SU downloads. First-time subscribing sites who are using NetDRMS 9.0 or later will automatically have automated downloads enabled by default. By default, sites that have recently (after the release of NetDRMS 9.0) become first-time subscribers will have these tasks performed automatically - the capture table, the capture function, and the capture triggers will all be created the first time a site registering for a subscription. Sites that had performed their first subscription before 9.0 was released lack the capture table, capture function, or one or more capture triggers, but those sites can manually create these objects: {{{ $ whoami <NetDRMS production user> $ <NetDRMS root>/base/drms/replication/subscribe_series/rsums-clientd.py [ --setup [ --capturesetup ] [ --seriessetup=<series> ]... ] }}} {{{--capturesetup}}} creates the capture table and capture function. {{{--seriessetup}}} creates the capture trigger for a single series. For example, to set up automatic SU downloads for hmi.M_45s and hmi.V_720s for a site that lacks the capture table and capture function, the {{{<NetDRMS production user>}}} can run: {{{ $ whoami <NetDRMS production user> $ rsums-clientd.py [ --setup [ --capturesetup ] [ --seriessetup=hmi.M_45s --seriessetup=hmi.V_720s ] ] }}} One all the mirror-request components are in place, the production user starts the Remote SUMS server (a daemon, rsumsd.py) if it is not already running. Then the user starts the mirror-request daemon (rsums-clientd.py). |
NetDRMS - a shared data management system
Introduction
In order to process, archive, and distribute the substantial quantity of solar data captured by the Atmospheric Imaging Assembly (AIA) and Helioseismic and Magnetic Imager (HMI) instruments on the Solar Dynamics Observatory (SDO), the Joint Science Operations Center (JSOC) has developed its own data-management system, NetDRMS. This system comprises two PostgreSQL databases, multiple file systems, a tape back-up system, and software to manage these components. Related sets of data are grouped into data series, each, conceptually, a table of data where each row of data typically associated with an observation time, or a Carrington rotation. As an example, the data series hmi.M_45s contains the HMI 45-second cadence magnetograms, both observation metadata and image FITS files. The columns contain metadata, such as the observation time, the ID of the camera used to acquire the data, the image rotation, etc. One column in this table contains an ID that refers to a set of data files, typically a set of FITS files that contain images.
The Data Record Management System (DRMS) is the subsystem that contains and manages the "DRMS" database of metadata and data-file-locator information. One component is a software library, written in C, that provides client programs, also known as "DRMS modules", with an Application Programming Interface (API) that allows the users to access these data. The Storage Unit Management System (SUMS) is the subsystem that contains and manages the "SUMS" database and associated storage hardware. The database contains information needed to locate data files that reside on hardware. The entire system as a whole is typically referred to as DRMS. The user interfaces with the DRMS subsystem only, and the DRMS subsystem interfaces with SUMS - the user does not interact with SUMS directly. The JSOC provides NetDRMS to non-JSOC institutions so that those sites can take advantage of the JSOC-developed software to manage large amounts of solar data.
A NetDRMS site is an institution with a local NetDRMS installation. It does not generate the JSOC-owned production data series (e.g., hmi.M_720s, aia.lev1) that Stanford generates for scientific use. A NetDRMS site can generate its own data, production or otherwise. That site can create software that uses NetDRMS to generate its own data series. But it can also act as a "mirror" for individual data series. When acting as a mirror for a Stanford data series, the site downloads from Stanford DRMS database information and stores it in its own NetDRMS database, and it downloads SUMS files, and stores them in its own SUMS subsystem. As the data files are downloaded to the local SUMS, the SUMS database is updated with the information needed to manage the data files. It is possible for a NetDRMS site to mirror the DRMS data of any other NetDRMS site, but at this point, the only site whose data are currently mirrored is the Stanford JSOC.
Installing NetDRMS
Installing the NetDRMS system requires:
installing PostgreSQL [ Installing PostgreSQL ]
instantiating a PostgreSQL cluster for two databases (one for DRMS and one for SUMS) [ Initializing PostgreSQL ]
installing CFITSIO [ Installing CFITSIO]
installing the DBD::Pg Perl package [ Installing DBD::Pg ]
installing packages to the system Python 3, or installing a new distribution, like Anaconda [ Installing Python3 ]
installing openssl development packages [ Installing OpenSSL Development Packages ]
installing the NetDRMS software code tree, which includes code to create DRMS libraries and modules and SUMS libraries [ Installing NetDRMS ]
initializing SUMS storage such as hard drives or SSD drives [ Initializing SUMS Storage ]
creating DRMS user accounts [ Creating DRMS User Accounts]
running the SUMS daemon (which accepts and processes SUMS requests from DRMS clients) [ Running SUMS]
Optional steps include:
registering for JSOC-data-series subscriptions and running NetDRMS software to receive, in real time, data updates [ Registering for Subscriptions ]
running the Remote SUMS daemon (which accepts and processes requests for SUs that reside at other NetDRMSs) [ Running Remote SUMS ]
- installing JSOC-specific project code that is not part of the base NetDRMS installation; the JSOC maintains code to generate JSOC-owned data that is not generally of interest to NetDRMS sites, but sites are welcome to obtain downloads of that code. Doing so involves additional configuration to the base NetDRMS system.
- installing Slony PostgreSQL data-replication software to become a provider of your site's data
- installing a webserver that hosts several NetDRMS CGIs to allow web access to your data
- installing the Virtual Solar Observatory (VSO) software to become a VSO provider of data
For best results, and to facilitate debugging issues, please follow these steps in order.
Conventions
In this document, parameters to be determined by you the NetDRMS administrator are denoted with angled brackes, <a parameter>. For example, you will need to select a machine to host the PostgreSQL database system, and the name of that host is represented by <PostgreSQL host>. If you choose a host named netdrms_db, let's say, then functionally, you can substitute netdrms_db for <PostgreSQL host> throughout this document.
Part of the process of installing NetDRMS is creating a configuration file named config.local. That file contains numerous parameters, such as SUMS_USEMTSUMS. Through this document, those parameters are denoted with square brackets. As such, this document refers to the the parameter SUMS_USEMTSUMS as [SUMS_USEMTSUMS].
Installing PostgreSQL
PostgreSQL is a relational database management system. Data are stored primarily in relations (tables) of records that can be mapped to each other - given one or more records, you can query the database to find other records. These relations are organized on disk in a hierarchical fashion. At the top level are one or more database clusters. A cluster is simply a storage location on disk (i.e., directory). PostgreSQL manages the cluster's data files with a single process, or PostgreSQL instance. Various operations on the cluster will result in PostgreSQL forking new ephemeral child processes, but ultimately there is only one master/parent process per cluster.
Each cluster contains the data for one or more databases. Each cluster requires a fair amount of system memory, so it makes sense to install a single cluster on a single host. It does not make sense to make separate clusters, each holding one database; each cluster can efficiently support many databases, which are then fairly independent of each other. In terms of querying the databases are completely independent (i.e., a query on one database cannot involve relations in different databases). However, two databases in a single cluster do share the same disk directory, so there is not the same degree of independence at the OS/filesystem level. This may only matter if an administrator is operating directly on the files (performing backups, replication, creating standby systems, etc.).
To install PostgreSQL, select a host machine, <PostgreSQL host>, to act as the PostgreSQL database server. We recommend installing only PostgreSQL on this machine, given the large amount of memory and resources required for optimal PostgreSQL operation. We find a Fedora-based system, such as CentOS, to be a good choice, but please visit https://www.postgresql.org/docs for system requirements and other information germane to installation. The following instructions assume a Fedora-based Linux system such as CentOS (documentation for other distributions, such as Debian and openSUSE can be found online) and a bash shell.
Install the needed PostgreSQL server packages on <PostgreSQL host> by first visiting https://yum.postgresql.org/repopackages.php to locate and download the PostgreSQL "repo" rpm file appropriate for your OS and architecture. Each repo rpm contains a yum configuration file that can be used to install all supported PostgreSQL releases. You should install the latest version if possible (version 12, as of the time of this writing). Although you can use your browser to download the file, it might be easier to use Linux command-line tools:
$ curl -OL https://download.postgresql.org/pub/repos/yum/reporpms/EL-7-x86_64/pgdg-redhat-repo-latest.noarch.rpm
Install the yum repo configuration file (pgdg-redhat-all.repo) from the downloaded repo rpm file:
$ sudo rpm -i pgdg-redhat-repo-latest.noarch.rpm
This installs the repo configuration file to /etc/yum.repos.d/. Find the names of the PostgreSQL packages needed from the repository; the following assumes PostgreSQL 12, but should you want to install an older version, replace "12" with one of 94, 95, 96, 10, or 11:
$ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql[0-9]*\.' | cut -d '.' -f 1 postgresql12 $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*devel\.' | cut -d '.' -f 1 postgresql12-devel $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*contrib\.' | cut -d '.' -f 1 postgresql12-contrib $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*libs\.' | cut -d '.' -f 1 postgresql12-libs $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*plperl\.' | cut -d '.' -f 1 postgresql12-plperl $ yum list --disablerepo='*' --enablerepo=pgdg12 2>/dev/null | grep -Eo '^.*postgresql.*server\.' | cut -d '.' -f 1 postgresql12-server
Use yum to install all four packages:
$ sudo yum install <packages>
where <packages> are the package names determined in the previous step (postgresql12 postgresql12-contrib postgresql12-devel postgresql12-libs postgresql12-plperl postgresql12-server). The rpm package installation will have created the PostgreSQL superuser Linux account <PostgreSQL superuser> (i.e., postgres); <PostgreSQL superuser> will own the PostgreSQL database clusters and server processes that will be created in the following steps. To perform the next steps, you will need to become user <PostgreSQL superuser>:
$ sudo su - <PosgreSQL superuser>
Depending on where the package files are installed, you might need to add the PostgreSQL command to your PATH environment variable. To test this, run:
$ which initdb
If the initdb command cannot be found, then add the PostgreSQL binaries path to PATH. Find the path to the PostgreSQL installation:
$ rpm -ql postgresql12 <PostgreSQL install dir>/bin/clusterdb ...
In this example, <PostgreSQL install dir> is /usr/pgsql-12. Then add the binary path to PATH:
$ export PATH=/usr/pgsql-12/bin:$PATH
<PosgreSQL superuser> will be using the binaries in this directory, so it is a good idea to add the export command to .bashrc. As described above, create one database cluster for the two databases (one for DRMS data, and one for SUMS data):
$ whoami postgres $ initdb --locale=C -D <PostgreSQL cluster>
where <PostgreSQL cluster> should be /var/lib/pgsql/netdrms. Use this path, unless there is some good reason you cannot. initdb will initialize the cluster data directory (identified by the -D argument). This will result in the creation of template databases, configuration files, and other items.
The database cluster will contain two configuration files you need to edit: postgresql.conf and pghba.conf. Please refer to the PostgreSQL documentation to properly edit these files. Here are some brief suggestions:
postgresql.conf - for changes to take effect of parameters marked with a *, a restart of a running server instance is required (pg_ctl restart), otherwise changes will require a reload (pg_ctl reload)
listen_addresses* specifies the interface on which the postgres server processes will listen for incoming connections. You will need to ensure that connections can be made from all machines that will run DRMS modules (the modules connect to both the DRMS and SUMS databases), so change the default 'localhost' to '*', which causes the servers to listen on all interfaces:
listen_addresses = '*'
port* is the server port on which the server listens for connections.
port = <PostgreSQL port>
The default port is 5432, and unless there is a good reason use 5432 for <PostgreSQL port>
logging_collector* controls whether on not stdout and stderr are logged to a file in the database cluster (in the log or pg_log directory, depending on release). By default it is off - set it so on in each cluster.
logging_collector = on
log_rotation_size sets the maximum size, in kilobytes, of a log file. Set this to 0 to disable rotation, otherwise a new log will be created after the current one grows to some size.
log_rotation_size = 0
log_rotation_age set the maximum age, in minutes, of a log file. Set this to 1d (1 day) so that each day a new log file is created.
log_rotation_age = 1d
log_min_duration_statement is the amount of time, in milliseconds, a query must run before triggering a log entry. Set to this 1000 so that only long-running queries, over a second, will be logged.
log_min_duration_statement = 1000
shared_buffers* is the size of shared-memory buffers. For a server dedicated to a single database cluster, this should be about 25% of the total memory.
shared_buffers = 32GB
pg_hba.conf controls the methods by which client authentication is achieved (HBA stands for host-based authentication). It will likely take a little time to understand and properly edit this configuration file. If you are not familiar with networking concepts (such as subnets, name resolution, reverse name resolution, CIDR notation, IPv4 versus IPv6, network interfaces, etc.) then now is the time to become familiar.
This configuration file contains a set of columns that identify which user can access which database from which machines. It also defines the method by which authenticaton occurs. When a user attempts to connect to a database, the server transverses this list looking for the first row that matches. Once this row is identified, the user must authenticate - if authentication fails, the connection is rejected. The server does not attempt additional rows. For changes to take effect of any of the parameters in this file, a reload of a server instance is required (not a restart)
Here are the recommended entries:
# local superuser connections # TYPE DATABASE USER AUTH-METHOD local all all trust # this applies ONLY if the user is logged into the PG server AND they do not use the -h argument to psql host all all 127.0.0.1/8 trust # for -h localhost, if localhost resolves to an IPv4 address; also for -h 127.0.0.1 host all all ::1/128 trust # for -h localhost, if localhost resolves to an IPv6 address; also for -h ::1 # non-local superuser connections # TYPE DATABASE USER ADDRESS AUTH-METHOD host all postgres XXX.XXX.XXX.XXX/YY trust # non-superuser connections (which can be made from any non-server machines only) # TYPE DATABASE USER ADDRESS AUTH-METHOD host netdrms all XXX.XXX.XXX.XXX/YY md5 host netdrms_sums all XXX.XXX.XXX.XXX/YY md5
where the columns are defined as follows:TYPE - this column defines the type of socket connection made (Unix-domain, TCP/IP, the encryption used, etc.). local is only relevant to Unix-domain local connections from the database server host <PostgreSQL host> itself. Since only postgres will log into the database server, the first row above applies to the administrator only. host is only relevant to TCP/IP connections, regardless of the encryption status of the connection.
DATABASE - this column identifies the database to which the user has access. Whenever a user attempt to connect to the database server, they specify a database to access. That database must be in the DATABASE column. We recommend using netdrms for non-superusers, blocking such users from accessing all databases except the DRMS one. Conversely, we recommend using all for the superuser so they can access both the DRMS and SUMS databases (and any other that might exist).
NOTE: You are using the database name netdrms in pg_hba.conf, even though you have not actually created that database yet. This is OK; you will do so once you start the PostgreSQL cluster instance.USER - this column identifies which users can access the specified databases.
ADDRESS - this column identifies the host IP addresses (or host names, but do not use those) from which a connection is allowed. To specify a range of IP addresses, such as those on a subnet, use a CIDR address. This column should be empty for local connections.
AUTH-METHOD - this column identifies the type authentication to use. We recommend using either trust or md5. When trust is specified, PostreSQL will unconditionally accept the connection to the database specified in the row. If md5 is specified, then the user will be required to provide a password. If you follow the recommendations above, then for the local row, any user who can log into the database server can access any database in the cluster without any further authentication. Generally only a superuser will be able to log into the database server, so this choice makes sense. For non-local connections by postgres, the Linux PostreSQL superuser postgres can access any database on the server without further authentication. For the remaining non-local non-postgres connections, users will need to provide a password.
Should you need to edit either of these configuration files AFTER you have started the database instance (by running pg_ctl start, as described in the next section), you will need to either reload or restart the instance:
$ whoami postgres # reload $ pg_ctl reload -D <PostgreSQL cluster> # restart $ pg_ctl restart -D <PostgreSQL cluster>
Initializing PostgreSQL
You need to now initialize your PostgreSQL instance by creating the DRMS and SUMS databases, installing database-server languages, creating a schema, creating a relation. To accomplish this become <PostgreSQL superuser>; all steps in this section must be performed by the superuser:
$ sudo su - postgres
Start the database instance for the cluster you created:
$ whoami postgres $ pg_ctl start -D <PostgreSQL cluster>
You previous created <PostgreSQL cluster>, which will most likely be /var/lib/pgsql/netdrms. Ensure the configuration files you created work. This can be done by attempting to connect to the database server as <PostgreSQL superuser> with psql from <PostgreSQL host>:
$ whoami <PostgreSQL superuser> $ hostname <PostgreSQL host> $ psql -h <PostgreSQL host> -p <PostgreSQL port> psql (12.1) Type "help" for help. postgres=# \q $
The PostgreSQL installation resulted in the creation of the postgres database superuser, and since psql connects to the database as the database user with the same name as the Linux user running psql, you will be logged in as database user postgres. This is indicated by the postgres=# prompt (the hash refers to a superuser).
After you successfully see the superuser prompt, create the two databases:
$ whoami postgres # create the DRMS database $ createdb --locale C -E UTF8 -T template0 netdrms # create the SUMS database $ createdb --locale C -E UTF8 -T template0 netdrms_sums
Install the required database-server languages:
$ whoami postgres # create the PostgreSQL scripting language (versions <= 9.6) # no need to create the PostgreSQL scripting language (versions > 9.6) $ createlang plpgsql netdrms # create the "trusted" perl language (versions <= 9.6) createlang -h <PostgreSQL host> plperl netdrms # create the "trusted" perl language (versions > 9.6) $ psql -h <PostgreSQLhost> -p <PostgreSQL port> netdrms netdrms=# CREATE EXTENSION IF NOT EXISTS plperl; netdrms=# \q # create the "untrused" perl language (versions <= 9.6) $ createlang -h <PostgreSQL host> plperlu netdrms # create the "untrused" perl language (versions > 9.6) netdrms=# CREATE EXTENSION IF NOT EXISTS plperlu; netdrms=# \q
The SUMS database does not use any language extensions so there is no need to create any for the SUMS database.
At this point, it is a good idea to create a password for the postgres database superuser:
$ whoami <PostgreSQL superuser> $ psql -h <PostgreSQLhost> -p <PostgreSQL port> netdrms netdrms=# ALTER ROLE postgres WITH PASSWORD '<new password>'; ALTER ROLE netdrms=# \q $
Installing CFITSIO
The base NetDRMS release requires CFITSIO, a C library used by NetDRMS to read and write FITS files. Visit https://heasarc.gsfc.nasa.gov/fitsio/ to obtain the link to the CFITSIO source-code tarball. Create an installation directory <CFITSIO install dir> (such as /opt/cfitsio-X.XX, with a link from /opt/cfitsio to <CFITSIO install dir>), download the tarball, and extract the tarball into <CFITSIO install dir>:
$ sudo mkdir -p <CFITSIO install dir> $ ctrl+d
$ curl -OL 'http://heasarc.gsfc.nasa.gov/FTP/software/fitsio/c/cfitsio-X.XX.tar.gz' $ tar xvzf netdrms_X.X.tar.gz $ cd cfitsio-X.XX
Please read the README file for complete installation instructions. As a quick start, run:
$ ./configure --prefix=<CFITSIO install dir> # build the CFITSIO library $ make # install the libraries and binaries to <CFITSIO install dir> $ sudo make install # create the link from cfitsio to <CFITSIO install dir> $ sudo su - $ cd <CFITSIO install dir>/.. $ ln -s <CFITSIO install dir> cfitsio
CFITSIO has a dependency on libcurl - in fact, any program made by linking to cfitsio will also require libcurl-devel since cfitsio uses the libcurl API. We recommend using yum to install the two packages (if they are not already installed - it is quite likely that libcurl will already be installed):
$ sudo yum install libcurl-devel
Installing OpenSSL Development Packages
NetDRMS requires the OpenSSL Developer's API. If this API has not already been installed, do so now:
$ sudo yum install openssl-devel
Installing DBD::Pg
One step in the installation process will require running a perl script that accesses the PostgreSQL database. In order for this to work, you will need to ensure the DBD::Pg module has been installed. To check for installation, run:
$ perl -M'DBD::Pg'
If there is no error about not being able to locate the module, and the command simply hangs, then you are all set (enter ctrl-C to exit). If the module is not installed, and you are running the perl installed with the system, then run yum to identify the package:
$ yum list | grep -i 'dbd-pg' ... perl-DBD-Pg.x86_64 2.19.3-4.el7 base ...
then bringing to bear all your powers of divination, choose the correct package and install it:
$ sudo yum install 'perl-DBD-Pg'
If using a non-system perl, use the distro's installation method. If the distro does not have that module, or the distro installer does not work, as a final act of desperation use CPAN
$ sudo perl -MCPAN -e 'install DBD::Pg'
Installing Python3
NetDRMS requires that a number of python packages and modules be present that are not generally part of a system installation. In addition, many scripts require python3 and not python2. The easiest way to satisfy these eeds is to install a data-science-oriented python3 distribution, such as Anaconda. In that vein, install Anaconda into an appropriate installation directory such as /opt/anaconda3. To locate the Linux installer, visit https://docs.anaconda.com/anaconda/install/linux/:
$ curl -OL 'https://repo.anaconda.com/archive/Anaconda3-2019.10-Linux-x86_64.sh' $ sha256sum Anaconda3-2019.10-Linux-x86_64.sh 46d762284d252e51cd58a8ca6c8adc9da2eadc82c342927b2f66ed011d1d8b53 Anaconda3-2019.10-Linux-x86_64.sh $ sudo bash Anaconda3-2019.10-Linux-x86_64.sh
After some initial prompts, the installer will display
PREFIX=/home/<user>/anaconda3
This path is the default installation directory (<user> is the user running bash). Replace the PREFIX path with <Anaconda3 install dir>.
Installing NetDRMS
To install NetDRMS, you will need to select an appropriate machine on which to install NetDRMS, an appropriate machine/hardware on which to host the SUMS service, create Linux users and groups, download the NetDRMS release tarball and extract the release source, initialize the Linux environment, create log directories, create the configuration file and run the configuration script, compile and install the executables, create the the DRMS- and SUMS-database users/relations/functions/objects, initialize the SUMS storage hardware, install the SUMS and Remote SUMS daemons.
The optimal hardware configuration will likely depend on your needs, but the following recommendations should suffice for most sites. DRMS and SUMS can share a single host machine. The most widely used and tested Linux distributions are Fedora-based, and at the time of this writing, CentOS is the most popular. Sites have successfully used openSUSE too, but if possible, we would recommend using CentOS. SUMS requires a large amount of storage to hold the DRMS data-series data/image files. The amount needed can vary widely, and depends directly on the amount of data you wish to keep online at any given time. Most NetDRMS sites mirror some amount of (but not all) JSOC SDO data - the more data mirrored, the larger the amount of storage needed. To complicate matters, a site can also mirror only a subset of each data series' data; perhaps one site wishes to retain only the current month's data of many data series, but another wishes to retain all data for one or two series. To decide on the amount of storage needed, you will have to ask the JSOC how much data each series comprises and decide how much of that data you want to keep online. Data that goes offline can always be retrieved automatically from the JSOC again. Data will arrive each day, so request from the JSOC an estimate of the rate of data growth. We recommend doing a rough calculation based upon these considerations, and then doubling the resulting number and installing that amount of storage.
Next, create a production Linux user <NetDRMS production user> (named netdrms_production by default):
$ sudo useradd <NetDRMS production user> $ sudo passwd <NetDRMS production user> Changing password for user <NetDRMS production user>. New password: Retype new password: passwd: all authentication tokens updated successfully. $
NOTE: ensure that <NetDRMS production user> is a valid PostgreSQL name because NetDRMS makes use of the PostgreSQL feature whereby attempts to connect to a database are made as the database user whose name matches the name of the Linux user connecting to the database. Please see https://www.postgresql.org/docs/12/sql-syntax-lexical.html#SQL-SYNTAX-IDENTIFIERS for a description of valid PostgreSQL names.
NetDRMS requires additional Python packages not included in the Anaconda distribution, but if you install Anacoda, then the number of additional packages you need to install is minimal. If you have a different python distribution, then you may need to install additional packages. To install new Anaconda packages, as <NetDRMS production user> first create a virtual environment for NetDRMS (named netdrms):
$ whoami <NetDRMS production user> # Sets path to Anaconda3 binaries so that the next conda command will success (will edit .bashrc) $ <Anaconda install dir>/bin/conda init bash $ source ~/.bashrc $ conda create --name netdrms Collecting package metadata (current_repodata.json): done Solving environment: done ## Package Plan ## environment location: /home/netdrms_production/.conda/envs/netdrms Proceed ([y]/n)? y Preparing transaction: done Verifying transaction: done Executing transaction: done # # To activate this environment, use # # $ conda activate netdrms # # To deactivate an active environment, use # # $ conda deactivate
VERY IMPORTANT!! As described above in the conda create output, run:
$ conda activate netdrms
If you do not do this, then conda might invoke the system Python, which could be version 2.7, when in fact your Anaconda is Python 3. If that happens, then all hope is lost and you'll need to remove the netdrms virtual environment and re-create it.
You will use conda to install additional Anaconda3 packages. To succees, you must have already installed PostgreSQL, and the PostgreSQL executables must be in $PATH (one step in the installation process is running pg_config. Add the following to <NetDRMS production user>'s .bashrc ensure PostgreSQL executables can be located:
$ whoami <NetDRMS production user> $ vi ~/.bashrc i ... # PostgreSQL executables export PATH=<PostgreSQL install dir>/bin:$PATH ESC :wq $
Make sure you have either re-logged-in or sourced .bashrc.
Then install the new Anaconda packages using conda:
$ whoami <NetDRMS production user> $ conda install -n netdrms psycopg2 psutil python-dateutil Collecting package metadata (current_repodata.json): done ...
NOTE: By installing these packages in the netdrms virtual environment, you will also install the python package in the environment since the explicitly listed packages have a dependency on python. This is important, because the next step requires that pip, part of python, be present in the environment.
The NetDRMS script needed for series subscription requires installation of the pySmartDL Python 3 package - a package that generally is not part of any distro. Unfortunately, conda supports this package for Python 2.7 only, and NetDRMS requires Python 3. But pip supports a Python 3 version. To install Python 3 pySmartDL, first make sure you have activated your virtual environment, as described above. Then run:
$ whoami <NetDRMS production user> $ which pip <NetDRMS production user home dir>/.conda/envs/netdrms/bin/pip $ pip install pySmartDL
If which pip displays an Anaconda path in <Anaconda3 install dir>, then do not run pip just yet. Go back and activate your virtual environment first.
From now on, the <NetDRMS production user> should use the virtual environment. You should make sure that $PYTHONPATH is not set, otherwise it might interfere with the running of Anaconda3 python. Modify that user's .bashrc to do so:
$ whoami <NetDRMS production user> $ vi ~/.bashrc i # .bashrc # use Python virtual environment by default unset PYTHONPATH # call conda activate AFTER # <<< conda initialize <<< conda activate netdrms ESC :wq $
For the changes to .bashrc to take effect, either logout/login or source .bashrc.
All non-production users will use the netdrms virtual environment. Make sure that users can access the environment:
$ whoami <NetDRMS production user> $ chmod o+x <NetDRMS production user home dir>
Create the Linux group <SUMS users>, e.g. sums_users, to which all SUMS users belong, including <NetDRMS production user>. This group will be used to ensure that all SUMS users can create new data files in SUMS:
$ sudo groupadd <SUMS users>
Add <NetDRMS production user> to this group (later you will add each SUMS user - users who will read/write SUMS data files - to this group as well):
$ sudo usermod -a -G <SUMS users> <NetDRMS production user> $ id <NetDRMS production user> uid=1001(netdrms_production) gid=1001(netdrms_production) groups=1001(netdrms_production),1002(sums_users)
On NetDRMS host, select a path <NetDRMS root> for the root directory of the NetDRMS source tree installed by <NetDRMS production user>. A typical choice for <NetDRMS root> is /opt/netdrms or /usr/local/netdrms, and a typical strategy is to install the source tree into a directory that contains the release version in its name (<NetDRMS install dir>), e.g., /opt/netdrms-9.3. Then <NetDRMS production user> makes a link from <NetDRMS root> to <NetDRMS install dir>. This facilitates the maintenance of multiple releases. To switch between releases, <NetDRMS production user> simply updates the link to point to the desired release directory. Create <NetDRMS install dir> and make <NetDRMS production user> the owner:
$ sudo mkdir -p <NetDRMS install dir> $ sudo chown <NetDRMS production user>:<NetDRMS production user> <NetDRMS install dir>
As <NetDRMS production user>, obtain a NetDRMS tarball from http://jsoc.stanford.edu/netdrms/dist/ and extract it into a release-specific directory:
$ whoami <NetDRMS production user> $ cd <NetDRMS install dir> $ curl -OL 'http://jsoc.stanford.edu/netdrms/dist/netdrms_X.X.tar.gz' $ tar xvzf netdrms_X.X.tar.gz $ <ctrl-d>
Create the link from <NetDRMS root> to <NetDRMS install dir>:
$ cd <NetDRMS install dir>/.. $ sudo ln -s <NetDRMS install dir> netdrms $
As <NetDRMS production user>, set the two environment variables that are needed for proper NetDRMS operation. To do so, you'll need first to determine the appropriate <architecture> string for one of these variables:
$ whoami <NetDRMS production user> $ cd <NetDRMS root> $ build/jsoc_machine.csh <architecture>
It is best to set the following two environment variables in <NetDRMS production user>'s .bashrc file since they must always be set whenever any NetDRMS code is run. And please put the NetDRMS executables in the path:
$ whoami <NetDRMS production user> $ vi ~/.bashrc i # .bashrc export JSOCROOT=/opt/netdrms export JSOC_MACHINE=<architecture> export PATH=/opt/netdrms/bin/<architecture>:$PATH ESC :wq $
Make the SUMS log directory <SUMS logs> on the SUMS server machine. Various SUMS log files will be written to this directory. A suitable directory would reside in the <NetDRMS production user> user's home directory, e.g., $HOME/log/SUMS
$ whoami <NetDRMS production user> $ mkdir -p <SUMS logs>
Select appropriate C and Fortran compilers. The DRMS part of NetDRMS must be compiled with a C compiler. NetDRMS supports both the GNU C compiler (gcc), and the Intel C++ compiler (icc). Certain JSOC-specific code requires Fortran compilation. For those projects, NetDRMS supports the GNU Fortran compiler (gfortran), and the Intel Fortran compiler (ifort). SUMS is implemented as a Python daemon, so no compilation step is needed. Both GNU and Intel are widely used, so feel free to use either. By default, Intel compilers are used. There are two methods for changing the compilers:
as the <NetDRMS production user>, you can set the JSOC_COMPILER and JSOC_FCOMPILER environment variables (we recommend doing so in .bashrc):
$ whoami <NetDRMS production user> $ vi i # .bashrc # set COMPILER to icc for the Intel C++ compiler, and to gcc for the GNU C++ compiler export JSOC_COMPILER=<C compiler> # set to ifort for the Intel Fortran compiler, and to gfortran for the GNU Fortran compiler export JSOC_FCOMPILER=<Fortran compiler> ESC :wq $
you can edit the make_basic.mk file in <NetDRMS root>; to select the Intel compilers, edit the COMPILER and FCOMPILER make variables declared near the top of the file:
# use Intel compilers COMPILER = icc FCOMPILER = ifort # use GNU compilers COMPILER = gcc FCOMPILER = gfortran
Create the <NetDRMS root>/config.local, the master configuration file for both DRMS and SUMS, using <NetDRMS root>/config.local.newstyle.template as a template. This file contains a number of configuration parameters, along with detailed descriptions of what they control and suggested values for those parameters. The configuration script, configure, reads this file, and then creates one output file, drmsparams.*, in <NetDRMS root>/localization for each of several programming languages/tools (C, GNU make, perl, python, bash). In this manner, the parameters are directly readable by several languages/tools used by NetDRMS. Lines that start with whitespace or the hash symbol, # are ignored.
NOTE: if you have an older NetDRMS and you are going to use the config.local from that NetDRMS installation for a new NetDRMS installation, then you might have an old-style config.local. You know you have an old-style configuration file if the __STYLE__ section does not exist in the file, or if it does exist, the value in that section is old. If that is the case, then you will need to compare <NetDRMS root>/config.local.template in the old installation to the analogous file in the new installation to determine the set of parameters that have changed between releases.
Several sections compose config.local:
__STYLE__ new __DEFS__ # these are NetDRMS-wide parameter values; the format is <quote code>:<parameter name><whitespace>+<parameter value>; # the configuration script uses <quote code> to assist in creating language-specific parameters; <quote code> is one of: # 'q' (enclose the parameter value in double quotes) # 'p' (enclose the parameter value in parentheses) # 'a' (do not modify the parameter value). __MAKE__ # these are make variables used by the make system during compilation - they generally contain paths to third-party code
Before creating config.local, please request from the JSOC a value for DRMS_LOCAL_SITE_CODE. This code uniquely identifies each NetDRMS installation. Each site requires one ID for each of its NetDRMS installations.
The __DEFS__ section:
BIN_PY3 - the path to the Python 3 python executable.
DBNAME - the name of the DRMS database: this is netdrms; this parameter exists in case you want to select a different name, but we don't recommend changing it.
DRMS_LOCAL_SITE_CODE - a 15-bit hexadecimal string that globally and uniquely identifies the NetDRMS. Each NetDRMS requires a unique code for each installation. Values greater than or equal to 0x4000 denote a development installation and need not be unique. If you plan on generating data that will be distributed outside of your site, please obtain a unique value from the JSOC.
DRMS_LOCK_DIR - the directory to which the DRMS library writes various lock files.
DRMS_LOG_DIR - the directory to which the DRMS library writes various log files.
EXPORT_HANDLE_DIR - the directory to which export programs save handles.
EXPORT_LOCK_DIR - the directory to which export programs write lock files.
EXPORT_LOG_DIR - the directory to which export programs write logs.
JMD_IS_INSTALLED - if set to 1, then the Java Mirroring Daemon alternative to Remote SUMS is used: this should be 0.
POSTGRES_ADMIN - the Linux user that owns the PostgreSQL installation and processes: this is <PostgreSQL superuser>.
RS_BINPATH - the NetDRMS binary path that contains the external programs needed by the Remote SUMS (e.g., jsoc_fetch, vso_sum_alloc, vso_sum_put).
RS_DBHOST - the name of the Remote SUMS database cluster host; this is <PostgreSQL host>, the machine on which PostgreSQL was installed.
RS_DBNAME - the Remote SUMS database - this is netdrms_remotesums.
RS_DBPORT - the port that the Remote SUMS database cluster instance is listening on: this is <PostgreSQL port>
RS_DBUSER - the Linux user that runs Remote SUMS; this is also the database user who owns the Remote SUMS database objects: this is <NetDRMS production user>.
RS_DLTIMEOUT - the timeout, in seconds, for an SU to download. If the download time exceeds this value, then all requests waiting for the SU to download will fail.
RS_LOCKFILE - the (advisory) lockfile used by Remote SUMS to prevent multiple instances from running.
RS_LOGDIR - the directory in which remote-sums log files are written.
RS_MAXTHREADS - the maximum number of SUs that Remote SUMS can process simultaneously.
RS_N_WORKERS - the number of scp worker threads - at most, this many scp processes will run simultaneously
RS_REQTIMEOUT - the timeout, in seconds, for a new SU request to be accepted for processing by the daemon. If the daemon encounters a request older than this value, it will reject the new request.
RS_REQUEST_TABLE - the Remote SUMS database relation that contains Remote SUMS requests; this is <Remote SUMS requests>; DRMS modules insert request rows in this table, and Remote SUMS locates the requests and manages rows in this table.
RS_SCP_MAXPAYLOAD - the maximum total payload, in MB, per download. As soon as the combined payload of SUs ready for download exceeds this value, then the SUs are downloaded with a single scp process.
RS_SCP_MAXSUS - the maximum size of the SU download queue. As soon as this many SUs are ready for download, they are downloaded with a single scp process.
RS_SCP_TIMEOUT - if there are SUs ready for download, and no scp has fired off within this many seconds, then the SUs that are ready to download are downloaded with a single scp process.
RS_SITE_INFO_URL - the service at JSOC that is used by Remote SUMS to locate the NetDRMS site that owns SUMS storage units; this is Remote SUMS site URL.
RS_SU_EXPIRATION - the default expiration date for all SUs ingested by Remote SUMS; if the SU being ingested is part of a data series, then Remote SUMS obtains the expiration for the SU from the data series' definition instead; as an alternative to RS_SU_EXPIRATION, RS_SU_LIFESPAN can be used to specify the expiration date of newly ingested SUs; RS_SU_EXPIRATION takes predent over RS_SU_LIFESPAN.
RS_SU_ARCHIVE - the default value of the archive flag for newly ingested SUs; if the SU being ingested is part of a data series, then Remote SUMS obtains the archive flag from the data series' definition instead; the truth value can be one of several character strings that implies TRUE or FALSE.
RS_SU_LIFESPAN - the default lifespan ("retention time"), in days, of a newly ingested SU; if the SU being ingested is part of a data series, then Remote SUMS obtains the lifespan for the SU from the data series' definition instead; as an alternative to RS_SU_LIFESPAN, RS_SU_EXPIRATION can be used to specify the lifespan of newly ingested SUs; RS_SU_EXPIRATION takes predent over RS_SU_LIFESPAN.
RS_SU_TAPEGROUP - the default value of the tapegroup for newly ingested SUs; if the SU being ingested is part of a data series, then Remote SUMS obtains the tapegroup from the data series' definition instead.
RS_TMPDIR - the temporary directory into which SUs are downloaded. This should be on the same file system on which the SUMS partitions reside.
SCRIPTS_EXPORT - the path to the directory in the NetDRMS installation that contains the export scripts.
SERVER - the name of the DRMS database cluster host: this is <PostgreSQL host>, the machine on which PostgreSQL was installed.
SS_HIGH_WATER - partition scrubbing is initiated only after partition percent usage rises above the high-water mark.
SS_LOCKFILE - the (advisory) lockfile used by the SU steward to prevent multiple instances of the steward from running.
SS_LOW_WATER - each SUMS partition is scrubbed until its percent usage falls below the low-water mark.
SS_REHYDRATE_INTERVAL - the time interval, in seconds, between updates to the per-partition cache of expired SUs; this value applies to all partitions that are scrubbed; for each partition, a steward thread queries its cache to select the next SUs to delete (which are sorted by increasing expiration date).
SS_SLEEP_INTERVAL - the interval, in seconds, between flushing/caching expired SU lists (use a smaller number if the system experience a high rate of SU expiration).
SS_SU_CHUNK - the number of SUs in a partition that are deleted at one time; SUs are deleted one chunk at a time until the partition usage falls below the low-water mark.
SUMBIN_BASEDIR - the directory in which sum_chmown, a root setuid program is installed; must be mounted locally on the machine on which the SUMS partition are mounted: this is <NetDRMS root>.
SUMLOG_BASEDIR - the path to the directory that contains various SUMS log files; this is <SUMS logs>.
SUMPGPORT - the port that the SUMS database cluster host is listening on: this is <PostgreSQL port>, unless DRMS and SUMS reside in different clusters on the same host (something that is not recommended since a single PostgreSQL cluster requires a substantial amount of system resources).
SUMS_DB_HOST - the name of the SUMS database cluster host: this is <PostgreSQL host>, the machine on which PostgreSQL was installed; NetDRMS allows for creating a second cluster for SUMS, but in general this will not be necessary unless extremely heavy usage requires separating the two clusters.
SUMS_GROUP - the name of the Linux group to which all SUMS Linux users belong: this is <SUMS users>.
SUMS_MANAGER - the SUMS database user who owns the SUMS database objects which are manipulated by Remote SUMS and SUMS itself; it should be the Linux user that runs SUMS and owns the SUMS storage directories - this is <NetDRMS production user>
SUMS_MULTIPLE_PARTNSETS - SUMS has more than one partition set: more than likely, this is 0.
SUMS_READONLY_DB_USER - the SUMS database user who has read-only access to the SUMS database objects; it is used by the Remote SUMS client (rsums-clientd.py) to check for the presence of SUs before requesting they be downloaded.
SUMS_TAPE_AVAILABLE - SUMS has a tape-archive system.
SUMS_USEMTSUMS - use the multi-threaded Python SUMS: this is 1.
SUMS_USEMTSUMS_ALL - use the multi-threaded Python SUMS for all SUMS API methods; SUMS_USEMTSUMS_ALLOC, SUMS_USEMTSUMS_CONNECTION, SUMS_USEMTSUMS_DELETESUS, SUMS_USEMTSUMS_GET, SUMS_USEMTSUMS_INFO, and SUMS_USEMTSUMS_PUT are ignored: this is 1.
SUMS_USEMTSUMS_ALLOC - use the MT SUMS daemon for the SUM_alloc() and SUM_alloc2() API function.
SUMS_USEMTSUMS_CONNECTION - use the MT SUMS daemon for the SUM_open() and SUM_close() API functions.
SUMS_USEMTSUMS_DELETESUS - use the MT SUMS daemon for the SUM_delete_series() API function.
SUMS_USEMTSUMS_GET - use the MT SUMS daemon for the SUM_get() API function.
SUMS_USEMTSUMS_INFO - use the MT SUMS daemon for the SUM_infoArray() API function.
SUMS_USEMTSUMS_PUT - use the MT SUMS daemon for the SUM_put() API function.
SUMSD_LISTENPORT - the port that SUMS listens to for incoming requests.
SUMSD_MAX_THREADS - the maximum number of SUs that SUMS can process simultaneously.
SUMSERVER - the SUMS host machine; this is <SUMS host>.
WEB_DBUSER - the DRMS database user account that cgi programs access when they need to read from or write to database relations.
The __MAKE__ section:
CFITSIO_INCS - the path to the installed CFITSIO header files: this is <CFITSIO install dir>/include.
CFITSIO_LIB - the name of the CFITSIO library.
CFITSIO_LIBS - the path to the installed CFITSIO library files: this is <CFITSIO install dir>/lib.
POSTGRES_INCS - the path to the installed PostgreSQL header files: this is <PostgreSQL install dir>/include.
POSTGRES_LIB - the name of the PostgreSQL C API library (AKA libpq): this is always pq.
POSTGRES_LIBS - the path to the installed PostgreSQL library files: this is <PostgreSQL install dir>/lib.
When installing NetDRMS updates, copy the existing config.local to the new <NetDRMS install dir> and edit the copy as needed, using the new config.local.newstyle.template to obtain information about parameters new to the newer release. Many of the parameter values have been determined during the previous steps of the installation process.
Run the configuration script, configure, a csh shell script which is included in <NetDRMS install dir>.
$ whoami <NetDRMS production user> $ cd <NetDRMS install dir> $ ./configure
configure reads config.local and uses the contained parameters to configure many of the NetDRMS features. It creates several directories in <NetDRMS install dir>:
bin - a directory that contains links to all executables in the DRMS code tree
include - a directory that contains links to all the header files in the DRMS code tree
jsds - a directory that contains links to all JSOC Series Definition (JSD) files
lib - a directory that contains links to all libraries in the DRMS code tree
localization - this directory contains project-specific make files and various files that contain the processed parameter information from config.local
scripts - a directory that contains links to all script files in the DRMS code tree
_<architecture> - a directory that contains all compiled files (object files, symbol files, binaries) in subdirectories that mirror the paths of the paths where the corresponding source code resides. For example, the source code for the show_info program resides in <NetDRMS install dir>/base/util/apps/show_info.c, and the show_info binary is placed in <NetDRMS install dir>/_<architecture>/base/util/apps/show_info. <architecture> was determined in a previous installation step. There are two possible names for <architecture>: _linux_x86_64 and _linux_avx (for hosts that support Advanced Vector Extensions).
Compile DRMS. To make the DRMS part of NetDRMS, run:
$ whoami <NetDRMS production user> $ cd <NetDRMS install dir> $ make
As <PostgreSQL superuser>, create the DRMS database production user <DRMS DB production user>. Since PostgreSQL automatically attempts to use the Linux user name as the PostgreSQL user name when a connection attempt is made, use Linux user <NetDRMS production user> for database user <DRMS DB production user>. Create a password for this user, and also GRANT all database privileges to <DRMS DB production user> (this is necessary so that <DRMS DB production user> can create namespaces in the netdrms database):
$ whoami <PostgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# CREATE ROLE <DRMS DB production user> WITH LOGIN; postgres=# ALTER ROLE <DRMS DB production user> WITH PASSWORD '<new password>'; postgres=# GRANT ALL ON DATABASE netdrms TO <DRMS DB production user>; postgres=# \q $
As <NetDRMS production user>, create a .pgpass file. This file contains the PostgreSQL user account password, obviating the need to manually enter the database password each time a database connection attempt is made:
$ whoami <NetDRMS production user> $ cd $HOME $ vi .pgpass i <PostgreSQL host>:*:*:<DRMS DB production user>:<new password> ESC :wq $ chmod 0600 .pgpass
As <PostgreSQL superuser>, run an SQL script, and a perl script (which executes several SQL scripts), both included in the NetDRMS installation, to create the admin and drms schemas and their relations, the jsoc and sumsadmin database users, data types, and functions:
$ whoami <PostgreSQL superuser> # use psql to execute SQL script $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U postgres -f <NetDRMS install dir>/base/drms/scripts/NetDRMS.sql netdrms CREATE SCHEMA GRANT CREATE TABLE CREATE TABLE GRANT GRANT CREATE SCHEMA GRANT CREATE ROLE CREATE ROLE $ perl <NetDRMS install dir>/base/drms/scripts/createpgfuncs.pl netdrms
For more information about the purpose of these objects, read the comments in the NetDRMS.sql and createpgfuncs.pl.
Although we'd like for the <DRMS DB production user> to own all these database objects that were just created by these scripts (these objects are all specific to NetDRMS), we had to run the scripts as <PostgreSQL superuser>, so the superuser owns them at this point. This was necessary since <DRMS DB production user> lacks the elevated privileges necessary to create these database objects. Run the following, as <PostgreSQL superuser> to alter ownerships:
$ whoami <PostgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> netdrms << EOF ALTER SCHEMA admin OWNER TO <DRMS DB production user>; ALTER TABLE admin.ns OWNER TO <DRMS DB production user>; ALTER TABLE admin.sessionns OWNER TO <DRMS DB production user>; ALTER SCHEMA drms OWNER TO <DRMS DB production user>; ALTER TYPE drmskw OWNER TO <DRMS DB production user>; ALTER TYPE drmsseries OWNER TO <DRMS DB production user>; ALTER TYPE drmssession OWNER TO <DRMS DB production user>; ALTER TYPE drmssg OWNER TO <DRMS DB production user>; ALTER TYPE rep_item OWNER TO <DRMS DB production user>; EOF ALTER SCHEMA ALTER TABLE ALTER TABLE ALTER SCHEMA ALTER TYPE ALTER TYPE ALTER TYPE ALTER TYPE ALTER TYPE $
We recommend using the NetDRMS database user <DRMS DB production user> as the SUMS database user <SUMS DB production user>. However, feel free to create a new user if necessary. If the DRMS and SUMS databases reside in different clusters, then you will need to create the <SUMS DB production user>. Again, since PostgreSQL automatically attempts to use the Linux user name as the PostgreSQL user name when a connection attempt is made, use Linux user <NetDRMS production user> for database user <SUMS DB production user>. If you choose a <SUMS DB production user> that is not <NetDRMS production user>, then you will need to pass <SUMS DB production user> to both Remote SUMS and SUMS when starting them.
$ # DO THIS ONLY IF <SUMS DB production user> != <DRMS DB production user> OR IF <DRMS database cluster> != <SUMS database cluster> $ whoami <PosgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# CREATE ROLE <SUMS DB production user>; postgres=# \q $
In addition, you will need to create a SUMS database user that has read-only access to the SUMS database objects:
$ whoami <PosgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# CREATE ROLE <SUMS DB readonly user> WITH LOGIN; postgres=# ALTER ROLE <SUMS DB readonly user> WITH PASSWORD 'readonlyuser'; postgres=# \q
where <SUMS DB readonly user> is [SUMS_READONLY_DB_USER]. This database account is used by the Remote SUMS Client, a daemon used to manage the auto-download of SUs for subscriptions. Remote SUMS Client will be run by <NetDRMS production user>, so add the password to <NetDRMS production user>'s .pgpass file:
$ whoami <NetDRMS production user> $ cd $HOME $ vi .pgpass i <PostgreSQL host>:*:*:[SUMS_READONLY_DB_USER]:readonlyuser ESC :wq $ chmod 0600 .pgpass
If you created a new SUMS DB production user, add a password for this user. Ensure that you use the same password that you used for <DRMS DB production user> - you will use the same Linux user when connecting to either database, so the same .pgpass file will be used for authentication. As <PostgreSQL superuser>, run psql to add a password for this new database user:
$ # DO THIS ONLY IF <SUMS DB production user> != <DRMS DB production user> $ whoami <PostgreSQL superuser> $ psql -h <PostgreSQL host> -p <PostgreSQL port> postgres=# ALTER ROLE <DRMS DB production user> WITH PASSWORD '<DRMS DB production user password>'; postgres=# \q $
SUMS stores directory and file information in relations in the SUMS database. To create those relations and initialize tables, as <NetDRMS production user> run:
$ whoami <NetDRMS production user> $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U <SUMS DB production user> -f <NetDRMS root>/base/sums/scripts/postgres/create_sums_tables.sql netdrms_sums CREATE TABLE CREATE INDEX CREATE INDEX GRANT CREATE TABLE GRANT CREATE TABLE GRANT CREATE INDEX CREATE INDEX CREATE INDEX CREATE INDEX CREATE TABLE GRANT CREATE TABLE CREATE INDEX GRANT CREATE SEQUENCE GRANT CREATE SEQUENCE GRANT CREATE TABLE GRANT CREATE TABLE GRANT CREATE TABLE GRANT $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U <SUMS DB production user> netdrms_sums netdrms_sums=> ALTER SEQUENCE sum_ds_index_seq START <min val> RESTART <min val> MINVALUE <min val> MAXVALUE <max val>; ALTER SEQUENCE netdrms_sums=> \q $
where <min val> is <drms site code> << 48, and <max val> is <min val> + <maximum unsigned 48-bit integer> - 1, where <drms site code> is the value of the [DRMS_LOCAL_SITE_CODE], and <maximum unsigned 48-bit integer> is 248 (which is 281474976710656). For the JSOC (site code 0x0000), this ALTER SEQUENCE command looks like:
netdrms_sums=> ALTER SEQUENCE sum_ds_index_seq START 0 RESTART 0 MINVALUE 0 MAXVALUE 281474976710655;
You now have the base part of NetDRMS installed. This excludes projects that are private to other NetDRMSs, such as the JSOC. It is often the case that your team would like to use private code as part of a data-generation pipeline. For example, the JSOC maintains time-distance analysis code that is part of the JSOC DRMS code tree, but not part of the base NetDRMS installation. To obtain this code, you can install those projects into your existing NetDRMS by: 1. obtaining the projects from the its source (like the JSOC), and placing it in your <NetDRMS install dir>/proj directory; and 1. modifying your config.local to refer to new projects, running configure, and then running make
Initializing SUMS Storage
In addition to SUMS database relations, SUMS requires a file system on which SUMS maintains storage areas called SUMS partitions. A SUMS paritition is really just a directory that contains SUMS Storage Units (each of which is implemented as a subdirectory inside the SUMS partition).
New Storage
As <NetDRMS production user> create one or more partitions now - although we have had success making them as large as 60 TB, make 40 TB partitions. For example, if you plan on setting aside X TB of SUMS storage, then make approximately N = <total storage TB> / X 40 TB partitions. The partitions can reside on a file server and be mounted onto all machines that will use NetDRMS, but the following example simply creates directories on a single file system on <SUMS partition host>. First, make the root directory that contains the SUMS paritions (something like /opt/sums):
$ sudo mkdir <SUMS root> # allow SUMS users to write into SUMS $ sudo chown <NetDRMS production user>:<SUMS users> <SUMS root> # when a user writes a file into SUMS, make sure that the file's group owner is <SUMS users> $ sudo chmod g+s <SUMS root>
<SUMS users> is the Linux group that is allowed to write to SUMS. You created it in a previous step. Second, make the SUMS partitions:
$ whoami <NetDRMS production user> $ hostname <SUMS partition host> $ mkdir -p <SUMS root>/partition01 $ mkdir -p <SUMS root>/partition02 ... $ mkdir -p <SUMS root>/partitionN
Initialize the SUMS DB sum_partn_avail table with the names of these partitions. For each SUMS partition run the following:
$ whoami <NetDRMS production user> $ psql -h <PostgreSQL host> -p <PostgreSQL port> -U <SUMS DB production user> netdrms_sums netdrms_sums=> INSERT INTO sum_partn_avail (partn_name, total_bytes, avail_bytes, pds_set_num, pds_set_prime) VALUES ('<SUMS partition path>', <avail bytes>, <avail bytes>, 0, 0);
where <SUMS partition path> is the full path of the partition as seen from <SUMS host> (which is where the SUMS daemon will run) and <avail bytes> is some number less than the number of bytes in the directory (multiply the number of blocks in the directory by the number of bytes per block). The number does not matter, as long as it is not bigger than the total number of bytes available. SUMS will adjust this number as needed.
Existing Storage
It might be the case that you would like to have a new NetDRMS installation use existing SUMS partitions. To do this, you would enter the paths to the existing partitions into the sum_partn_avail table. At this point, the SUMS database tables that contain the pointers to SUs (sum_main and sum_partn_alloc) do not contain any references to the existing SUs. Most likely the existing SUMS partitions are part of an existing, old NetDRMS, in which case these references are in that old NetDRMS SUMS database table. If so, then you can manually dump four database tables/sequences from the old NetDRMS to files that can be ingested into the new NetDRMS installation. To do that, as the old NetDRMS <NetDRMS production user> run the following:
$ whoami <NetDRMS production user> $ psql -h <OLD NetDRMS host> -p <OLD NetDRMS port> netdrms_sums netdrms_sums=> COPY public.sum_main TO '/tmp/sum_main_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_partn_alloc TO '/tmp/sum_partn_alloc_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_seq TO '/tmp/sum_seq_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_ds_index_seq TO '/tmp/sum_ds_index_seq_dump.txt' WITH ENCODING 'UTF8';
The COPY command will save the table data onto a file on the OLD NetDRMS PostgreSQL server. Copy those files to the NEW NetDRMS <PostgreSQL host> and then ingest them:
$ whoami <NetDRMS production user> $ psql -h <NEW NetDRMS host> -p <NEW NetDRMS port> netdrms_sums netdrms_sums=> COPY public.sum_main FROM '/tmp/sum_main_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_partn_alloc FROM '/tmp/sum_partn_alloc_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_seq FROM '/tmp/sum_seq_dump.txt' WITH ENCODING 'UTF8'; netdrms_sums=> COPY public.sum_ds_index_seq FROM '/tmp/sum_ds_index_seq_dump.txt' WITH ENCODING 'UTF8';
Creating DRMS User Accounts
For each Linux user, <new DRMS user> who will run installed DRMS modules, or who will write new DRMS modules, you will need to set up their environment, create their DRMS account, and create their .pgpass file so they can run DRMS modules without having to manually authenticate to the database. Just like you did for the <NetDRMS production user>, you will need to unset the PYTHONPATH environment variable, and put PostgreSQL executables into the user's $PATH. You will also need to put the <NetDRMS production user>'s netdrms Python virtual environment into the user's $PATH - each user will use this Python environment, not the base installation. Also set the JSOCROOT, JSOC_MACHINE, COMPILER, and FCOMPILER environment variables:
# .bashrc # PostgreSQL executables export PATH=/usr/pgsql-12/bin:$PATH # python production virtual environment unset PYTHONPATH export PATH=<NetDRMS production home>/.conda/envs/netdrms/bin:$PATH # NetDRMS binary path export PATH=<NetDRMS root>/bin/<architecture>:$PATH # root of NetDRMS source/installation export JSOCROOT=<NetDRMS root> # architecture export JSOC_MACHINE=<architecture> # set COMPILER to icc for the Intel C++ compiler, and to gcc for the GNU C++ compiler export COMPILER=<C compiler> # set to ifort for the Intel Fortran compiler, and to gfortran for the GNU Fortran compiler export FCOMPILER=<Fortan compiler>
You will also need to add them to the <SUMS users> group:
$ sudo usermod -a -G <SUMS users> <new DRMS user>
To create a DRMS account, you create a database account for the user, plus you add user-specific rows to various DRMS database tables. The script newdrmsuser.pl exists to facilitate these tasks:
$ whoami <NetDRMS production user> $ perl <NetDRMS root>/base/drms/scripts/newdrmsuser.pl netdrms <PostgreSQL host> <PostgreSQL port> <new DRMS user> <initial password> <new DB user namespace> user 1 Connection to database with 'dbi:Pg:dbname=netdrms;host=drms;port=<PostgreSQL port>' as user '<new DRMS user>' ... success! executing db statment ==> CREATE USER <new DRMS user> executing db statment ==> ALTER USER <new DRMS user> WITH password '<initial password>' executing db statment ==> GRANT jsoc to <new DRMS user> running cmd-line ==> masterlists dbuser=<new DRMS user> namespace=<new DB user namespace> nsgrp=user Please type the password for database user "postgres": Connected to database 'netdrms' on host '<PostgreSQL host>' and port '<PostgreSQL port>' as user 'postgres'. Created new drms_series... Created new 'drms_keyword'... Created new 'drms_link'... Created new 'drms_segment'... Created new 'drms_session'... Created new drms_sessionid_seq sequence... Commiting... Done. executing db statment ==> INSERT INTO admin.sessionns VALUES ('<new DRMS user>', '<new DB user namespace>')
where <new DB user namespace> is the PostgreSQL namespace dedicated to the new user. A namespace is a logical container that allows a database user to own database objects, like relations, that have the same name as objects owned by other users - items in a namespace need only be uniquely named within the namespace, not between namespaces. For example, the relation drms_series in the namespace su_arta is not the same relation as the drms_series relation in the su_phil namespace - the relations have the same name, but they are different relations. In virutally all PostgreSQL operations, a user can prefix the name of a relation with the namespace: su_arta.drms_series refers to the first relation, and su_phil.drms_series refers to the second relation.
The purpose of <new DB user namespace> is to hold non-production, private data series - sort of a private user space to develop new DRMS modules to create data. If those data should become a production-level products, then the data and the code that generates the data need to be moved to a production namespace. At the JSOC, we have several such production namespaces (e.g., aia, hmi, mdi). A site creates production namespaces with a different module (masterlists; newdrmsuser.pl is only for creating non-production namespaces.
Please see the NOTE in this page for assistance with choosing <new DB user namespace>. The general naming convention is to prepend the namespace with an abbreviation to identify the site that owns the data in the namespace. For example, all private data created at Stanford reside in dataseries whose namespaces start with su_ (Stanford University), regardless of the affiliation of the user who creates data in this namespace. Data created at NASA Ames start with nas_ (NASA Supercomputing Division). Following the underscore is a string to identify a particular user - su_arta for Art, and su_phil for Phil. You can also specify a group with the suffix (e.g., su_uscsolar for a solar group at the University of Southern California that creates data at Stanford. <initial password> is the initial password for this account - the initial password does not matter much since you are going to have the user change it next.
Running newdrmsuser.pl will create a new DRMS database user that has the same name as the user's Linux account name.
Have the user change their password:
$ whoami <new DRMS user> $ psql -h <PostgreSQL host> -p <PostgreSQL port> netdrms netdrms=> ALTER USER <new DRMS user> WITH PASSWORD '<new password>'; netdrms=> \q $
And then have the user create their .pgpass file (to allow auto-login to their database account) and set permissions to 0600:
$ whoami <new DRMS user> $ cd $HOME $ vi .pgpass i <PostgreSQL host>:*:*:<DRMS DB production user>:<new password> ESC :wq $ chmod 0600 .pgpass
Please click here for additional information on the .pgpass file.
If you plan on creating data that will be publicly distributed, you should also create one or more data-production users. For example, if you plan on making a new public HMI data series, you could create a user named hmi_production. Although you could follow previous steps to create a new Linux account for this database user, you do not necessarily need to. Instead you can use the existing <NetDRMS production user> and have it connect as the hmi_production user. To do that, first create the new hmi_production database user by running newdrmsuser.pl as just described. Choose a descriptive namespace that follows the naming guidelines described about, like hmi. Because a .pgpass already exists for <NetDRMS production user>, you want to ADD a new line to .pgpass for this user. Continuing with the hmi_production user example, add a password line for hmi_production:
# .pgpass <PostgreSQL host>:*:*:hmi_production:<hmi_production password>
At this point, it is a good idea to test that the new DRMS user can use a basic NetDRMS program. Although your NetDRMS has no DRMS data series, running show_series is a good way to test various components, like authentication, database connection, etc. Test DRMS by running the show_series command:
$ whoami <new DRMS user> $ show_series $
Nothing will be printed when this command is run, since your NetDRMS is devoid of data at the moment, but if you see no errors, then life is good. If not, then contact the JSOC for help troubleshooting.
To perform a more thorough test involving SUMS, you will need to have a least one DRMS data series that has SUMS data. You can obtain such a data series by registering for subscriptions.
Running SUMS Services
Before you can use NetDRMS, you, as <NetDRMS production user> on <SUMS host>, will need to start SUMS. To launch one or more SUMS daemons, sumsd.py, use the start-mt-sums.py script:
$ whoami <NetDRMS production user> $ hostname <SUMS host> $ python3 <NetDRMS root>/base/sums/scripts/start-mt-sums.py daemon=<NetDRMS root>/base/sums/scripts/sumsd.py ports=6102 --loglevel=debug running /home/netdrms_production/.conda/envs/netdrms/bin/python3 /opt/netdrms/base/sums/scripts/sumsd.py --sockport=6102 --loglevel=debug started instance /opt/netdrms/base/sums/scripts/sumsd.py:6102 (pid 27515) {"started": [27515]
}
This command starts sumsd.py, and creates the instances file and a log file in SUMLOG_BASEDIR by default. The instances file is a state file used by start-mt-sums.py and stop-mt-sums.py to manage the running sumsd.py instances. By default, the log file is named sumsd-<PPPPP>-<YYYYMMDD.HHMMSS>.txt, where <PPPPP> is the PID of the sumsd.py process, and <YYYYMMDD.HHMMSS> is the time string representing the time the instance was launched.
The complete usage is:
usage: start-mt-sums.py daemon=<path to daemon> ports=<listening ports> [ --instancesfile=<instances file path> ] [ --loglevel=<critical, error, warning, info, or debug>] [ --logfile=<file name> ] [ --quiet ] optional arguments: -h, --help show this help message and exit -i <instances file path>, --instancesfile <instances file path> the json file which contains a list of all the sumsd.py instances running -l LOGLEVEL, --loglevel LOGLEVEL specifies the amount of logging to perform; in order of increasing verbosity: critical, error, warning, info, debug -L <file name>, --logfile <file name> the file to which sumsd logging is written -q, --quiet do not print any run information required arguments: d <path to daemon>, daemon <path to daemon> path of the sumsd.py daemon to launch p <listening ports>, ports <listening ports> a comma-separated list of listening-port numbers, one for each instance to be spawned
start-mt-sums.py will fork one or more sumsd.py daemon processes. The ports argument identifies the SUMS host ports on which sumsd.py will listen for client (DRMS module) requests. One sumsd.py process will be invoked per port specified. Each process creates a log-file in [SUMLOG_BASEDIR] named, by default, sumsd-<port>-<YYYYMMDD>.<HHMMSS>.txt. The -L/--logfile argument allows you to override the path and name to this log file.
To stop one or more SUMS services, use the stop-mt-sums.py script:
$ whoami <NetDRMS production user> $ hostname <SUMS host> $ python3 <NetDRMS root>/base/sums/scripts/stop-mt-sums.py daemon=<NetDRMS root>/base/sums/scripts/sumsd.py --ports=6102
The complete usage is:
usage: stop-mt-sums.py [ -h ] daemon=<path to daemon> [ ---ports=<listening ports> ] [ --instancesfile=<instances file path> ] [ --quiet ] optional arguments: -h, --help show this help message and exit -p <listening ports>, --ports <listening ports> a comma-separated list of listening-port numbers, one for each instance to be stopped -i <instances file path>, --instancesfile <instances file path> the json file which contains a list of all the sumsd.py instances running -q, --quiet do not print any run information required arguments: d <path to daemon>, daemon <path to daemon> path of the sumsd.py daemon to halt
Registering for Subscriptions
A NetDRMS site can optionally register for a data-series subscription to any NetDRMS site that offers subscription service. The JSOC NetDRMS offers subscriptions, but at the time of this writing, no other site does. Once a site registers for a data series subscription, the site will become a mirror for that data series. The subscription process ensures that the mirroring site will receive regular and timely updates made to the data series by the serving site. The subscribing site can configure the interval between updates such that the mirror can synchronize with the server and receive updates within a couple of minutes, keeping the mirror up-to-date in (almost) real time.
To register for a subscription, <Subscription production user> will set up ssh keys (for SU transfer), start daemons, and run the subscription script, subscribe.py. The assumption is that <Subscription production user> is <NetDRMS production user>, but you are free to choose a different user. subscribe.py makes subscription requests to the serving site's subscription manager. The process entails the creation of a snapshot of the data-series DRMS database information at the serving site. Those data are downloaded, via HTML, to the subscribing site, where they are ingested by subscribe.py. get_slony_logs.pl, a client-side cron task, updates the data-series snapshot with any server-side changes that have been made since the snapshot was created. Ingestion of the snapshot results in the creation of the DRMS database objects that maintain and store the data series, and get_slony_logs.pl updates those objects when the server makes changes. At this time, no SUMS data files are downloaded. Instead, and optionally, the IDs for the series' SUMS Storage Units (SU) are saved in a database relation. It is the function of Remote SUMS (rsums.py), another NetDRMS daemon, to download the SUs and ingest them into your SUMS.
Remote SUMS accepts requests from DRMS for SUs. It then communicates with the serving NetDRMS and manages the scp download and ingestion of those SUs. Once Remote SUMS is running, should any DRMS code/module request an SU that is not present in the local NetDRMS, DRMS will send a download request to Remote SUMS. The Remote SUMS Client (rsums-clientd.py), an optional NetDRMS daemon, can automate this process so that when new subscription data are ingested into the DRMS database, it submits requests for the associated SUs to Remote SUMS. In this way, it is possible to pre-fetch the SU files before any user requests them. But pre-fetching is optional. The SUs will be downloaded on-demand as described above. In fact, if the subscribing NetDRMS site were to automatically download an SU, then delete the SU (there is a method to do this, described later), then an on-demand download is the only way to re-fetch the deleted SU. On-demand downloads happen automatically; any DRMS module that attempts to access an SU (like with a show_info -p command) that is not present for any reason will trigger an rsumsd.py request. The module will pause until the SU has been downloaded, then automatically resume its operation on the previously missing SU.
As get_slony_logs.pl users scp to download update files (Slony logs) and rsumsd.py uses scp to automatically download SUs, SSH public-private keys must be created at the subscribing site, and the public key must be provided to the serving site. Setting this up requires coordinated work at both the susbscribing and serving sites. As <Subscription production user> on the subscribing site, run
$ whoami <Subscription production user> $ ssh-keygen -t rsa
This will allow you to create a passphrase for the key. If you choose to do this, then save this phrase for later steps. In the home directory of <Subscription production user>, ssh-keygen will create a public key named id_rsa.pub. Provide this public key to the serving site.
The serving site must then add the public key to its list of authorized keys. If the .ssh directory does not exist, then the serving site must first create this directory and give it 0700 permissions. If the authorized_keys file in .ssh does not exist, then it must first be created and given 0644 permissions:
$ whoami <Subscription manager user> $ mkdir .ssh $ chmod 0700 .ssh $ cd .ssh $ touch authorized_keys $ chmod 0644 authorized_keys
Once the .ssh and authorized_keys files exist and have the proper permissions, the serving site administrator can then add the client site's public key to its list of authorized keys:
$ whoami <Subscription manager user> $ cd $HOME/.ssh $ cat <remote-site public key file> >> authorized_keys
Back at the client NetDRMS site, if an ssh passphrase was chosen, then as <Subscription production user> start an ssh-agent instance to automate the passphrase authentication. If no passphrase was provided when ssh-keygen was run, this step can be skipped. Otherwise, run:
$ whoami <Subscription production user> $ ssh-agent > $HOME/.ssh-agent $ source $HOME/.ssh-agent # needed for ssh-add, and also for rsumsd.py and get_slony_logs.pl $ ssh-add $HOME/.ssh/id_rsa
and provide the passphrase. To keep ingested data series synchronized with changes made to it at the serving site, a client-side cron tab runs get_slony_logs.pl periodically. This perl script uses scp to download slony log files - SQL files that insert, delete, or update database relation rows. get_slony_logs.pl communicates with the Slony-I replication software running at the serving site. Slony-I generates these log (SQL) files at the server which are then downloaded by the client.
To register for a subscription to a new series, run:
$ python3 <NetDRMS root>/base/drms/replication/subscribe_series/subscribe.py cfg=<subscription config file> reqtype=subscribe series=<published DRMS data series> --loglevel=debug
The complete usage is:
usage: subscribe.py [ -hjpl ] cfg=<client configuration file> reqtype=<subscribe, resubscribe, unsubscribe> series=<comma-separated list of series> [ --archive=<0, 1>] [ --retention=<number of days>] [ --tapegroup=<tape-group number> ] [ --pg_user=<subscripton client DB user> ] [ --logfile=<log-file name> ] optional arguments: -h, --help show this help message and exit archive <series archive flag>, --archive <series archive flag> The tape archive flag for the series - either 0 (do not archive) or 1 (archive). retention <series SU disk retention>, --retention <series SU disk retention> The number of days the series SUs remain on disk before becoming subject to deletion. tapegroup <series SU tape group>, --tapegroup <series SU tape group> If the archive flag is 1, the number identifying the group of series that share tape files. pg_user <series SU tape group>, --pg_user <series SU tape group> The DB account the subscription client uses. -p, --pause Pause and ask for user confirmation before applying the downloaded SQL dump file. --loglevel LOGLEVEL Specifies the amount of logging to perform. In increasing order: critical, error, warning, info, debug --logfile <file name> The file to which logging is written. --filtersus <SU filter> Specifies a series keyword K and a number of days D, comma separated; a remote-SUMS request for an SU will occur only if the keyword K of the record containing the SU has a value that lies within the time interval determined by the days D. required arguments: cfg <client configuration file>, --config <client configuration file> The client-side configuration file used by the subscription service. reqtype <request type>, --reqtype <request type> The type of request (subscribe, resubscribe, or unsubscribe). series SERIES, --series SERIES A comma-separated list of DRMS series to subscribe/resubscribe to, or to unsubscribe from.
The debug level logging is not necessary, but if this is your first subscription, we recommend running with lots of debug statement - if there are issues, you can send this output to the JSOC for help.
<subscription config file> contains parameters used by both the subscription client subscribe.py and the program that downloads Slony logs file, get_slony_logs.pl. We recommend making a directory subscription in the home directory of <NetDRMS production user> and saving the configuration file in that directory. The parameters to be set in <subscription config file> are as follows:
node - the name of the subscription client (e.g., jsoc, nsocu, sdac; must be globally unique across all NetDRMS sites; this string will be used in various state files and in file/directory names.
kRSServer - the full domain name of the subscription log server (e.g., jsocport.stanford.edu for a client subscibing to data series published by the JSOC).
kRSUser - the account on kRSServer that will be used for data transfer (e.g., jsocexp for a client subscibing to data series published by the JSOC).
kRSPort - the port on kRSServer that will be used for data transfer (e.g., 22 for scp).
kRSBaseURL - the base URL for all subscription services provided by subscription server for the Slony cluster (identified by slony_cluster)
pg_host - the client machine that hosts the client PostgreSQL database that will contain the replicated data series - this is <PostgreSQL host>.
pg_port - the port on the pg_host machine that will be used for communication with the data-series database - this is <PostgreSQL port>.
pg_user - the PostgreSQL user that will own the replicated series - this is <NetDRMS production user>.
pg_dbname - the name of the PostgreSQL database that resides on pg_host - this is netdrms.
slony_cluster - the name of the Slony cluster to which this node belongs (e.g., jsoc for a client subscribing to data series published by the JSOC).
kLocalLogDir - the client directory that will contain the subscription-process logs; we recommend <NetDRMS production user home>/subscription/log.
kLocalWorkingDir - the path to a directory for temporary working subscription files
kSQLIngestionProgram - the path to the script/program that will ingest the site-specific slony logs — usually the path to get_slony_logs.pl (<NetDRMS root>/base/drms/replication/get_slony_logs.pl
kSubService - the URL of the application at the subscription-serving site that accepts new subscription requests (for the JSOC subscripton server this is ${kRSBaseURL}/request-subs.py)
kPubListService - the URL of the application at the subscription-serving site that that lists published data series (for the JSOC subscripton server this is ${kRSBaseURL}/publist.py)
kSubXfer - the URL of the application at the subscription-serving site where subscription dump files are located (for the JSOC subscripton server this is http://jsoc.stanford.edu/subscription)
kDeleteSeriesProgram - the path to the program delete_series, which is used to delete DRMS data series on the client when requested (<NetDRMS root>/bin/<architecture>/delete_series).
archive - for new subscriptions, the default data series archive action; set to 1 if the NetDRMS site has a tape archive system AND the default is to archive all series obtained by subscription
retention - for new subscriptions, the default number of days to retain SUs (after this many days SUs are marked for deletion and subjected to garbage collection as needed)
tapegroup - for new subscriptions, the default archive tape group for the the series' SUs (ignored if archive == 0)
ingestion_path - the local directory that will contain the ingestion "die" file — used by get_slony_logs.pl
scp_cmd - the absolute path to the client's scp program.
ssh_cmd - the absolute path to the client's ssh program.
rmt_slony_dir - the absolute path, accessible from the kRSUser account, on the server to the directory that contains the site-specific slony logs (/data/pgsql/slon_logs/live/site_logs).
slony_logs - the client directory that contains the downloaded site-specific slony logs.
PSQL - the path to the client's psql program, and any flags needed to run psql as the pg_user user, like -h pg_host.
email_list - the email account to which error messages will be sent.
You may find that a subscription has gotten out of sync, for various reasons, with the serving site's data series (accidental deletion of database rows, for example). You can re-register for the subscription to true-up. The existing DRMS data will be deleted, and replaced with a fresh snapshot. You will be prompted asking if you would like to delete the series' SUMS data (SUs, FITS files, etc.). If you are sure you no longer need them, go ahead and say yes.
python <NetDRMS root>/base/drms/replication/subscribe_series/subscribe.py cfg=<subscription config file> reqtype=resubscribe series=<subscription series> --loglevel=debug
Finally, there might come a time where you no longer which to hold on to a registration. To remove the subscription from your set of registered data series run: subscribe.py can be used to alleviate this problem. Run the following to re-do the subscription registration:
python <NetDRMS root>/base/drms/replication/subscribe_series/subscribe.py cfg=<subscription config file> reqtype=unsubscribe series=<subscription series> --loglevel=debug
NOTE: In this case, the subscribe.py will prompt you to ask if you would like to delete the existing DRMS data series. The answer is usually yes. However, if you need to keep the existing data series snapshot for some reason (e.g., you want to work with the existing data, but simply do not want to ingest new data), then respond no. Keep in mind that you will not be able to register for a subscription to the same series if that series already exists - you will need to run delete_series.
In all the above commands, the logging level is set to debug. It is a good idea to enable verbose logging like this the first time you run one of these commands, just in case an issue occurs (usually do to some configuration issue). Providing the debug log to the JSOC when troubleshooting will be invaluable.
Once you have successfully registered for subscription to at least one series, you will need to install a crontab to run get_slony_logs.pl, which updates the subscription series, on a regular basis. get_slony_logs.pl has a dependency on the Net::SSH perl package. If you are running CentOS, then there is a package available to install this package:
If you cannot find a package, then you can use CPAN.
Once you have the Net::SSH Perl package installed, run subscribe.py as <Subscription production user>:
$ whoami <Subscription production user> $ crontab -e */5 * * * * (. ~/.ssh-agent; <NetDRMS root>/base/drms/replication/get_slony_logs.pl <subscription config file> >> ~/<path>/get_slony_logs.log 2>&1 )
At the JSOC, Slony log files are created every minute. It is not the case that your series' subscriptions will have been updated every minute (this depends on the cadence of the series to which you have subscriptions), but checking every minute will minimize the lag between the state of your series and the state of your series at the serving site.
Running Remote SUMS
Your NetDRMS may contain data produced by other, non-local NetDRMSs. Via a variety of means, the local NetDRMS can obtain and ingest the database information for these data series produced non-locally. In order to use the associated data files (typically image files), the local NetDRMS must download the storage units (SUs) associated with these data series too. Remote SUMS, a tool that comes with NetDRMS, downloads SUs as needed - i.e., if a DRMS module or program requests the path to the SU or attempts to read it, and it is not present in the local SUMS yet, Remote SUMS will download the SUs. While the SUs are being downloaded, the initiating module or program will poll waiting for the download to complete.
Several components compose Remote SUMS. On the client side, the local NetDRMS, is a daemon that must be running rsumsd.py. There also must exist some database tables, as well as some binaries used by the daemon. On the server side, all NetDRMS sites that wish to act as a source of SUs for the client, is a CGI (rs.sh). This CGI returns file-server information (hostname, port, user, SU paths, etc.) for the SUs the server has available in response to requests that contain a list of SUNUMs. When the client encounters requests for remote SUs that are not contained in the local SUMS, it sends a request the Remote SUMS daemon to download those SUs. The client code then polls waiting for the request to be serviced. The daemon in turn sends requests to all rs.sh CGIs at all the relevant providing sites. The owning sites return the file-server information to the daemon, and then the daemon downloads the SUs the client has requested, via scp, and notifies the client module once the SUs are available for use. The client module will then exit from its polling code and continue to use the freshly downloaded SUs.
To use Remote SUMS, several config.local parameters must be present. If you followed the steps in this document, then you have already set those parameters. Please see Installing NetDRMS for a description of each one. Each SU that is downloaded has an associated expiration date, a flag indicating whether or not the SU is archived, and if the SU is archived the tapegroup to which the SU belongs. The manner in which the values for these parameters are determined is a bit complicated. When you register for a series subscription, the series is created at your site, and at that time, the values for these parameters are determined. The series is then initialized with these values. There are default value for these parameters that are overridden by optional parameters supplied during the call to subscribe.py. In the following list, the lower-numbered items, if present, override the higher-numbered items, in this order:
the --archive, --retention, --tapegroup command-line arguments to subscribe.py
the archive, retention, tapegroup parameters in subscription config file
Now, when Remote SUMS runs, the values of the parameters for SUs downloaded and ingested are determined in a similar hierarchical fashion, with higher-numbered items overriding lower-numbered items:
- the parameters associated with the series, as determined above
the --archive, --expiration, --lifespan, --tapegroup command-line arguments to rsumsd.py
[RS_SU_ARCHIVE], [RS_SU_EXPIRATION], [RS_SU_LIFESPAN], [RS_SU_TAPEGROUP]
To run Remote SUMS, as NetDRMS production user run the following to create the requests database table:
$ whoami <NetDRMS production user> $ python3 <NetDRMS root>/base/drms/scripts/rscreatetabs.py tabs=req
Remote SUMS downloads SUs via scp. As such, you will need to create SSH keys, distribute the public one to the site serving the SUs, and start up an ssh-agent if you have not already done so - you should have already though since it is needed by get_slony_logs.pl, a component of the subscription system (please see [ Registering for Subscriptions ]).
To launch Remote SUMS, ensure sumsd.py is first running, then as <NetDRMS production user> source the ssh-agent environment file, and then run:
$ whoami <NetDRMS production user> $ source $HOME/.ssh-agent $ python <NetDRMS root>/base/drms/scripts/rsumsd.py --loglevel=debug
For now, we recommend setting the log level to debug; when things appear to be running smoothly, then you can restart with the default level (info). The output log named <rslog_YYYYMMDD.txt> will be written to the directory identified by [RS_LOGDIR], so make sure that directory exists before running rsumsd.py.
Running Remote SUMS Client
If you have at least one subscription, and you have Remote SUMS running (not the JMD), then you can automate the download of SUs for the subscriptions. This is optional, however. If you skip this step, then when a NetDRMS user attempts to use one or more SUs that are part of a subscription (and hence not at your NetDRMS site initially), the SUs will be downloaded in an on-demand fashion. However, it may be desirable to pre-fetch the SUs if certain usage patterns hold. If one or more users are going to use a block of SUs for a two-week period, for example, then downloading hundreds or thousands of SUs one at a time would be very inefficient. Also, it is very common for NetDRMS users to use newly produced SUs soon after they are created at the subscription server. In this case, a good strategy might be to automatically download the latest SUs for all subscriptions, knowing that lots of very recent SUs will be popular.
The Remote SUMS Client, rsums-clientd.py, monitors SU references in incoming Slony logs. It groups SUs into batches, and submits requests containing these batches to Remote SUMS ,rsumsd.py. rsums-clientd.py then monitors the progress of these requests, logging the results.
In addition to a running rsumsd.py, rsums-clientd.py requires the existence of three components. The capture table is a database table that contains a list of all SUNUMs that are to be downloaded. It gets automatically populated as Slony logs are ingested. The capture function is a database function that performs the actual work of inserting rows into the capture table. It is called every time a Slony log inserts a row into a subscribed series' series table. You will need to create a capture trigger database trigger for each series table for which you want to automate SU downloads. These components work together as follows:
each capture trigger watches a series table, and when a row is inserted (due to get_slony_logs.pl ingesting new Slony logs), the trigger runs the capture function
- this function then inserts a row into the capture table
rsums-clientd.py then sees the newly inserted capture-table row and extracts the SUNUM
rsums-clientd.py batches several of these SUNUMs, and then makes a rsumsd.py request out of them
rsumsd.py processes the requests (multiple in parallel), downloading one or more SUs and ingesting them into the local SUMS
rsumsd.py updates a status column in the capture table
rsums-clientd.py reads the status, and upon success, it removes the rows that contained the SUs that were sucessfully ingested; upon failure, rsums-clientd.py will log an error, and can optionally re-try one or more times
To use rsums-clientd.py, <NetDRMS production user> must first create the capture table and the capture function. They must also add the capture trigger to the series table of the DRMS data series for which they want to enable automatic SU downloads. First-time subscribing sites who are using NetDRMS 9.0 or later will automatically have automated downloads enabled by default.
By default, sites that have recently (after the release of NetDRMS 9.0) become first-time subscribers will have these tasks performed automatically - the capture table, the capture function, and the capture triggers will all be created the first time a site registering for a subscription. Sites that had performed their first subscription before 9.0 was released lack the capture table, capture function, or one or more capture triggers, but those sites can manually create these objects:
$ whoami <NetDRMS production user> $ <NetDRMS root>/base/drms/replication/subscribe_series/rsums-clientd.py [ --setup [ --capturesetup ] [ --seriessetup=<series> ]... ]
--capturesetup creates the capture table and capture function. --seriessetup creates the capture trigger for a single series.
For example, to set up automatic SU downloads for hmi.M_45s and hmi.V_720s for a site that lacks the capture table and capture function, the <NetDRMS production user> can run:
$ whoami <NetDRMS production user> $ rsums-clientd.py [ --setup [ --capturesetup ] [ --seriessetup=hmi.M_45s --seriessetup=hmi.V_720s ] ]
One all the mirror-request components are in place, the production user starts the Remote SUMS server (a daemon, rsumsd.py) if it is not already running. Then the user starts the mirror-request daemon (rsums-clientd.py).