IRMIS

IRMIS Overview

"IRMIS is a collaborative effort among several EPICS sites to build a common Relational DataBase schema and a set of tools to populate and search an RDB that contains information about the operational EPICS IOCs installed at that site." IRMIS (the schema, crawler programs and UI) was developed by Don Dohan and Claude Saunders at APS. For general information and distributions see the IRMIS home page.

IRMIS @ SLAC

IRMIS is used at SLAC for the following purposes, from several different interfaces:

• Data source for PV names for AIDA (nightly cron jobs)
• Data source for element EPICS device names for LCLS_INFRASTRUCTURE (nightly cron jobs)
• PV and PV Client lists and data (IRMIS gui)
• IOC configuration and parameters (IOC Info APEX: https://seal.slac.stanford.edu/apex/mccqa/f?p=104:8)
• IOC and application configuration data (IOC Info jsp app: https://seal.slac.stanford.edu/IRMISQueries )
• EPICS camdmp application (APEX app: https://seal.slac.stanford.edu/apex/mccqa/f?p=103:4)
• Archiver PV search application (APEX app: https://seal.slac.stanford.edu/apex/mccqa/f?p=259:8 )
• Lists of IOCs and their PV populations (web page: http://mccas0.slac.stanford.edu/crawler/ioc_report.html)
• ad hoc queries using TOAD, sqlplus, pl/sql, perl scripts, or other db query tool.
• IRMIS crawler logs, with duplicate PV reports: http://www.slac.stanford.edu/grp/lcls/controls/sysGroup/report

Elements of the IRMIS database that have been adopted and modified for the controls software group at SLAC:

• The Oracle PV schema
• PV, ALH crawlers
• the IRMIS gui
• Other elements of the collaboration IRMIS installation include cabling, device and application schemas. We are not populating these now...but may in the future.

Elements modified or created at SLAC:

• IOC boot syntax adaptations in the PV crawler and IOC Info crawler (adopted from SNS) and schema
• Special application-oriented tables in the IRMIS schema, and their population scripts (for example for EPICS camdmp and the archiver PV viewer app)
• Addition of config file tree structure crawling in the ALH crawler
• New crawlers for Channel Watcher and Channel Archiver
• New PV Client viewer addition to the IRMIS Desktop
• New guis

#top

UI and DATABASE QUERYING

IRMIS GUI
This is a java UI for the IRMIS Oracle database, developed by Claude Saunders of the EPICS collaboration, can be invoked in 3 ways:

1. from the lclshome edm display:
   click the “IRMIS…” button
2. from a Solaris or Linux workstation:
   run this script:
   irmisUI

The gui paradigm is a set of “document types”; click the File/New Document menu for the list. Right now there are only 2 available for use at SLAC:

1. idt::pv – Search for lists of PVs and IOCs. This is the most useful interface, and it comes up upon application startup.
2. idt::pvClient – Search for PV Client lists (alarm handler, archiver, channel watcher)

Query results can be saved to an ascii file for further processing.

IOC Parameters APEX application
This is an APEX application showing IOC configuration data, and various operational parameter snapshots (obtained live nightly using caget).
https://seal.slac.stanford.edu/apex/mccqa/f?p=104:8

EPICS camdmp APEX application
Various reports listing PVs and their module and channel connections.
https://seal.slac.stanford.edu/apex/mccqa/f?p=103:4

Archiver PV query APEX application
Lists of archived PVs by IOC application, IOC.
https://seal.slac.stanford.edu/apex/mccqa/f?p=259:8

IOC list report (run nightly)

This html report is created nightly by the LCLS PV Crawler
http://mccas0.slac.stanford.edu/crawler/ioc_report.html

PV Crawler logs (which include the duplicate PV lists)
http://www.slac.stanford.edu/grp/lcls/controls/sysGroup/report/
a subset here: http://www.slac.stanford.edu/cgi-bin/lwgate/CONTROLS-SOFTWARE-REPORTS/archives

SQL querying
A view has been created to ease sql querying for PV lists. This view combines data from the IOC_BOOT, IOC, REC and REC_TYPE tables. It selects currently loaded PVs, where IOC_BOOT.CURRENT_LOAD = 1, with the latest IOC boot date captured.

• CURR_PVS view has all currently loaded PVs

Discontinued: IOC Info query application
This is a jsp web application containing data about IOCs and applications, and their configurations.  It is no longer maintained.
https://seal.slac.stanford.edu/IRMISQueries/

#top

OPERATIONAL and SUPPORT DETAILS

At SLAC, in a nutshell the PV and PV client crawlers update:

• all IRMIS tables in MCCQA. All IRMIS UIs query data on MCCQA: PV and IOC data, and PV Client data.
• 3 production tables on MCCO, which are used by other system – AIDA, BSA applications

Oracle schemas and accounts

The IRMIS database schema is installed in 4 SLAC Oracle instances:

• MCCQA IRMISDB schema: Production for all gui’s and applications. Contains data populated nightly by perl crawler scripts from production IOC configuration files. Data validation is done following each load.
• MCCO IRMISDB schema: contains production data, currently for 3 tables only: bsa_root_names, devices_and_attributes, curr_pvs. MCCQA data is copied to MCCO once it has been validated. So MCCO is as close as possible to pristine data at all times.
• SLACDEVcontains the 3 schemas which are used for developing, testing and staging new features before release to production on MCCQA (see accounts below)
  1. IRMISDB – sandbox for all kinds of development, data sifting, etc. Not refreshed from prod, or not very often.
  2. IRMISDB_TEST – testing for application implementation. Refreshed from prod at the start of a development project.
  3. IRMISDB_STAGE – staging for testing completed applications against recently-refreshed production data.
• SLACPROD IRMISDB schema: this is obsolete, but will be kept around for awhile (6 months?) as a starting point in case the DB migration goes awry somehow.

Other Oracle accounts

• IRMIS_RO – read-only account (not used much yet – but available)
• IOC_MGMT – created for earlier IOC info project with a member of the EPICS group which is not active at the moment- new ioc info work is being done using IRMISDB

For passwords see Judy Rock or Poonam Pandey or Elie Grunhaus.

As of September 15, all crawler-related shell scripts and perl scripts use Oracle Wallet to get the latest Oracle password. Oracle passwords must be changed every 6 months, and is coordinated and implemented by Poonam Pandey.

Database structure

see schema diagram below. (this diagram excludes the EPICS camdmp structure, which is documented separately here: <url will be supplied>)

Crawler scripts

The PV crawler is run once for each IOC boot directory structure. The LCLS and FACET PV crawlers are separate to enable them to be run on a different schedule and different host which can see the production IOC directories for these systems. Also, the LCLS crawler code has been modified to be LCLS-specific; it is a different version than the SLAC PV crawler.

The SLAC PV crawler (runSLACPVCrawler.csh) runs the crawler a couple of times to accommodate the various CD IOC directory structures.

cron jobs

• LCLS side: laci on lcls-daemon2: runLCLSPVcrawlerLx.bash: crawls LCLS PVs and creates lcls-specific tables (bsa_root_names, devices_and_attributes), copies LCLS client config files to dir where CD client crawlers can see them.
• LCLS side: laci on lcls-daemon2: caget4curr_ioc_device.bash: does cagets to populate curr_ioc_devices for the IOC Info APEX app. Run separately from the crawlers because cagets can hang unexpectedly – they are best done in an isolated script!
• CD side: laci on lcls-prod01: runAllCDCrawlers.csh: runs CD PV crawler and all client crawlers, data validation, and sync to MCCO.
• FACET side: flaci on facet-daemon1: runFACETPVcrawlerLx.bash: crawls FACET PVs, copies FACET client config files to dir where CD client crawlers can see them.

PV crawler operation summary

For the location of the crawler scripts, see Source code directories below.
Basic steps as called by cron scripts are:

1. run FACET pv crawler to populate MCCQA tables
2. run LCLS pv crawlers to populate MCCQA tables
3. run CD pv and pv client crawlers to populate MCCQA tables
4. run Data Validation for PV data in MCCQA
5. if Data Validation returns SUCCESS, run synchonization of MCCQA data to selected (3 only at the moment) MCCO tables.
6. run caget4curr_ioc_device to populate caget columns of curr_ioc_device

** For PV Crawlers: the crawler group for any given IOC is determined by its row in the IOC table. The system column refers to the boot group for the IOC, as shown below.

* The PV client crawlers load all client directories in their config files; currently includes both CD and LCLS.

LOGFILES, Oracle audit table

Log filenames are created by appending a timestamp to the root name shown in the tables below.

The major steps in the crawler jobs write entries into the Oracle CONTROLS_GLOBAL.DATA_VALIDATION_AUDIT table. Each entry has these attributes:

• Instance
• Schema
• Process
• Stage
• Status
• Message
• TOD (time of day)
  (see below for details on querying this table)

Descriptions of the MAIN scripts (there other subsidiary scripts as well):

these are all ultimately invoked from the cron jobs shown above; the cron scripts call the others.

BLUE script names are on the CD side

GREEN script names are on the LCLS side

PURPLE script names are on the FACET side

#top

Crawler cron schedule and error reporting

Following the crawls, the calling scripts grep for errors and warnings, and send lists of these to Judy Rock, Bob Hall and Ernest Williams, Jingchen Zhou. The LCLS PV crawler and the Data Validation scripts send messages to controls-software-reports as well. To track down the error messages in the e-mail, refer to the logs file du jour, the cron job /tmp output files, and the Oracle CONTROLS_GLOBAL.DATA_VALIDATION_AUDIT (see below for details on how to query this)

Source directories

#top

PV Crawler: what gets crawled and how

• runSLACPVCrawler.csh, runLCLSPVCrawlerLx.bash and runFACETPVCrawlerLx.bash set up for and run the IRMIS pv crawler multiple times to hit all the boot structures and crawl groups. Environment variables set in pvCrawlerSetup.bash, pvCrawlerSetup-facet.bash and pvCrawlerSetup.csh, point the crawler to IOC boot directories, log directories, etc. Throughout operation, errors and warnings are written to log files.
• The IOC table in the IRMIS schema contains the master list of IOCs. The SYSTEM column designates which crawler group the IOC belongs to.
• An IOC will be hit by the pv crawler if its ACTIVE column is 1 in the IOC table: ACTIVE is set to 0 and 1 by the crawler scripts, depending on SYSTEM, to control what is crawled. i.e. LCLS IOCs are set to active by the LCLS crawler job. NLCTA IOCs are set to active by the CD crawler job.
• An IOC will not be crawled unless its LCLS/FACET STARTTOD or CD TIMEOF BOOT (boot time) PV has changed since the last crawl.
• Crawler results will not be saved to the DB unless at least one file mod date or size has changed.
• Crawling specific files can be triggered by changing the date (e.g. touch).
• In the pv_crawler.pl script, there’s a mechanism for forcing the crawling of subsets of IOCs (see the code for details)
• IOCs without a STARTTOD/TIMEOFBOOT PV will be crawled every time. (altho PV data is only written when IOC .db files have changed)

#top

TROUBLESHOOTING: post-crawl emails and data_validation_audit entries: what to worry about, how to bypass or fix for a few…

• The DATA_VALIDATION FINISH step MSG field in the CONTROLS_GLOBAL.DATA_VALIDATION_AUDIT table will show the list of errors found by the data validation proc:
  • Log into Oracle on MCCQA as IRMISDB
    • use TOAD or
    • from the Linux command line:
      • source /usr/local/lcls/epics/setup/oracleSetup.bash
      • export TWO_TASK=MCCQA
      • sqlplus irmisdb/`getPwd irmisdb`

select * from controls_global.data_validation_audit where schema_nm=’IRMISDB’ order by tod desc;

This will show you (most recent first) the data validation entries, with any error messages.

You can also see a complete listing of data_validation_audit entries in reverse chron order by using AIDA (but you will have to pick out the IRMISDB lines):

• Launch AIDA web https://mccas1.slac.stanford.edu/aidaweb
• In the query line, enter LCLS//DBValidationStatus
  ** Please focus only on entries where schema_nm is IRMISDB. This report shows entries for ALL of our database operations; sometimes entries from different systems are interleaved.*

For a “good” set of IRMISDB entries, scroll down and see those from: 9/22 9 pm, continuing into 9/23, which completed successfully. The steps in ascending order are:

FACET PV Crawler start
FACET PV Crawler finish
ALL_DATA_UPDATE start
LCLS PV Crawler start
LCLS PV Crawler finish
CD PV Crawler start
CD PV Crawler finish
PV Client Crawlers start
PV Client Crawlers finish
PV Client Cleanup start
PV Client Cleanup finish
DATA_VALIDATION start
DATA_VALIDATION finish
FACET DATA_VALIDATION start
FACET DATA_VALIDATION finish
ALL_DATA_UPDATE finish

then there are multiple steps for the sync to MCCO, labelled REFRESH_MCCO_IRMIS_TABLES

some specific error checks:

• comparison of PV counts: pre and post crawl
  An error is flagged if the total PV count has dropped more than 5000, because this may indicate that some IOCs have been skipped that were previously being crawled. This would happen when a change occurs in the IOC boot files or IOC boot structure that the crawler doesn’t understand. For example, if the crawler encounters new IOC startup file syntax that it doesn’t understand, but that doesn’t actually crash the crawler program.
  The logfile and e-mail message will tell you which IOCs are affected. The first thing to do is check with the responsible IOC engineer(s). It’s possible that the PV count drop is “real”, i.e. the engineer(s) in question intentionally removed a large block of PVs from the system.

• • If the drop is intentional: You will need to update the database manually to enable the crawler to proceed: the count difference is discovered by comparing the row count of the newly populated curr_pvs view with the row count of the materialized view curr_pvs_mv, which was updated on the previous day. If the PV count has dropped > 5000 PVs, the synchronization is cautious - it prevents updating good data in MCCO until the reason for the drop is known.
    The step to say "it's ok, the drop was intentional or at least non-destructive - synchronization is now ok" is to go ahead and updated the materialized view with current data. Then the next time the crawler runs, the counts will be closer (barring some other problem) and the synchronization can proceed
    So, to enable the crawlers proceed, you need to update the materialized view with data in the current curr_pvs view, like this:
    log in as laci on lcls-prod01 and run
    /afs/slac/g/cd/tools/irmis/cd_script/updateMaterializedViews.csh
    The next time the crawler runs (that night), the data validation will be making the comparison with correct current data, and you should be good to go.

• • If the drop is NOT intentional:Check the log for that IOC (search for “Processing iocname”) and/or ask the IOC engineer to check for changes in the IOC boot directories, and IOC boot files for:
    • • a syntax error in the boot file or directory
      • new IOC st.cmd syntax has change to something that the crawler hasn’t learned yet. May require a PVCrawlerParser.pm modification to teach IRMIS a new syntax. This is rare but does occur. It’s possible the IOC engineer can switch back to an older-style syntax, add quotes around a string, etc. to temporarily adjust the situation.

• • comparison of processed IOC counts in 2 log files
    If the numbers don't match, the PV Crawler terminated early, before it processed all the IOCs. This usually happens when the crawler encounters new IOC startup file syntax that crashes the crawler perl program. Go to the logfile to determine the cause. This is rare, but usually requires a PVCrawlerParser.pm modification. As a temporary fix, update the IOC table and for the offending IOC(s), set SYSTEM to something other than LCLS or NLCTA (e.g. LCLS-TEMP-NOCRAWL) so it/they will not be crawled next time.

• • bsa_root_names row count checks: check for problem accessing the table, or 0 or otherwise unexpected row counts for expected bsa types

• • devices_and_attributes row count checks: problem accessing the table, or row count 0, of devices_and_attributes row count differs from distinct rows in the curr_pvs view.
  • If there is a discrepancy:
    • A possible candidate is PVs with the same name in both FACET and LCLS, or between more than 1 IOC in LCLS.  Since devices_and_attributes has unique PV names, while in curr_pvs PV name+IOC name, names duplicated in curr_pvs will only show up once in devices_and_attributes, hence a count difference.  Here is a useful select statement to find the duplicates:
    • select distinct (rec_nm) from curr_pvs where system='LCLS'
      minus
      select distinct(rec_nm) from devices_and_attributes where system='LCLS'
    • The fix will have to occur in the implicated IOC application...and the next crawl following the fix will resolved the problem.
    • For a shortterm mitigation, edit IRMISDataValidation.pl and comment out the check_DEVICES_AND_ATTRIBUTES step.

• • check for successful completion of all crawler steps: check to make sure all crawler steps enter a start and a completion row in controls_global.data_validation_audit.
    some circumstances where steps are missing:
    • step launched but didn’t finish: check the status of processes launched by the cron job using ps --ef | grep. An example: when perl dbi was hanging due to the 199-day-Linux-server-uptime bug. Several LCLS PV crawler jobs had launched, but had hung in the db_connect statement, and had to be killed from the Linux command line.
    • step launched and finished, but the completed step was never written: the getPwd problems cause this symptom. See entries starting 9/23 9 pm for an illustration.
    • step never launched: is the script available? Is the server up? Is crontab/trscrontab configured correctly? Are there permission problems? etc.
    • other mysteries: figure out where the job in question stopped, using ps --ef, logfiles, etc…

Logfile messages sent in e-mail:

• • Duplicate PV report (lcls only near the top of the message)
    Forward these to the relevant IOC Engineer.
  • List of IOCs with PV counts that changed during the current crawl (lcls only near the top of the message) – helpful for figuring out a big PV count drop
  • IOCS not found in IRMIS, added to IOC table (lcls only) – nice to know which new IOCs were just added.
    If an IOC directory in $IOC is not in the IRMIS IOC database table yet, IRMIS will automatically add it, and produce a message which you'll see here.
  • PV Crawler log messages
    • Don't worry about "expecting 3 arguments to nfsMount"!
    • Quickly check "Could not locate" - usually an IOC dir that isn't quite in production yet.
    • You can send a message to the IOC engineer if you see " unreachable" - it could be a typo, or it could be intentional.
    • A "Could not locate" an st.cmd file will create "Parse error: 2" messages, which can be ignored if the IOC isn't in production.

To bypass crawling specific IOC(s): update the IOC table and for the IOC(s) in question, set SYSTEM to something other than LCLS or NLCTA (e.g. LCLS-TEMP-NOCRAWL) so it/they will not be crawled next time.

#top

In the event of an error that stops crawling, here are the affected end-users of IRMIS PVs:

(usually not an emergency, just decay of the accuracy of the lists over time.)

• CURR_PVS: IRMIS PVs create a daily current PV list, a view called curr_pvs. curr_pvs supplies lcls pv names to one of the AIDA names load jobs (LCLS EPICS names). The current PV and IOC lists are also queried by the IRMIS gui, and by web interfaces, and joined with data in lcls_infrastructure by Elie and co. for Greg and co.
• BSA_ROOT_NAMES: Qualifying IRMIS PVs populate the bsa_root_names table, which is joined in with Elie's complex views to device data in lcls_infrastructure. For details on the bsa_root_names load, please have a look at the code for the stored procedure which loads it (see one of the tables above)
• DEVICES_AND_ATTRIBUTES:* PV names are parsed into device names to populate the devices_and_attributes table which is used by lcls_infrastructure and associated processing.
• User interfaces may be affected (but they use MCCQA, so are not affected by failure of the sync to MCCO step):
  • IRMIS gui
  • the web IOC Report
  • IRMIS iocInfo web and APEX apps
  • EPICS camdmp APEX app
  • Archiver PV search APEX app
  • ad hoc querying

#top

3 possible worst case workarounds:

1. Synchronization to MCCO that bypasses error checking
   • If you need to run the synchronization to MCCO even though IRMISDataValidation.pl failed (i.e. the LCLS crawler ran fine, but others failed), you can run a special version that bypasses the error checking, and runs the sync no matter what. It’s:
     /afs/slac/u/cd/jrock/jrock2/DBTEST/tools/irmis/cd_script/runSync-no-check.csh
2. Comment out code in IRMISDataValidation.pl
   • If the data validation needs to bypass a step, you can edit IRMISDataValidation.pl (see above tables for location) to remove or change a data validation step and enable the crawler jobs to complete. For example, if a problem with the PV client crawlers causes the sync to MCCO not to run, you may want to simply remove the PV Client crawler check from the data validation step.
3. Really worst case! Edit the MCCO tables manually
   • If the PV crawlers will not complete with a sync of good data to MCCO, and you decide to wait til November for me to fix it (this is fine – the PV crawler parser is a complicated piece of code that needs tender loving care and testing!), AND accelerator operations are affected by missing PVs in one of these tables, the tables can be updated manually with names that are needed to operate the machine:
     • aida_names (see Bob and Greg)
     • bsa_root_names (see Elie)
     • devices_and_attributes (see Elie)

#top

crawling a brand new directory structure

• start off by testing in the SLACDEV instance!! To do this, must check the pv crawler dir out of cvs into a test directory, and modify db.properties to point to SLACDEV instance PLUS cvs checkout crawl scripts and setenv TWO_TASK to SLACDEV for set*IOCsActive scripts.
• all directories have to be visible from each the IOC boot directory
• the host running the crawler must be able to "see" Oracle (software and database) and the boot directory structure.
• set up env vars in run*PVCrawler.csh and pvCrawlerSetup.csh.
• if necessary, create crawl group in IOC table, and set*IOCsActive.csh to activate.
• add call to pv_crawler.csh to run with the new env vars...
• hope it works!

#top

SCHEMA DIAGRAM