GCS software development

Tuesday 26 May 2020, 15:00 → 18:00 Europe/Zurich

12/05/2020

Alarm/warning monitoring:

statistics: add screenshot of plots for every meeting -> Mara

NA62 Straw 1 warning to tune -> Roberto (ongoing)

LHC vs non-LHC egroup for alarms/warnings (Michal to create ticket and add ref.number here): GMXR-32 ,  GNA-96

egroup for warnings to be added for LINAC, CMSmixers and NA62

We keep the full stop egroup for NA62.

Any upgrade of hardware or functionalities for non-LHC systems?

LINAC, NA62 use Siemens PLC: any plan to upgrade? -> Michal to check with BE

touch screen for LINAC B152. How much does it cost? -> Michal to check - 2000 to 3000e depend on the size - new touch screen will also require development of the touch screen application (touch screens recommended by BE-ICS: https://readthedocs.web.cern.ch/pages/viewpage.action?pageId=106594570)

Link to summary: https://espace.cern.ch/gas-control/_layouts/15/start.aspx#/Shared%20Documents/Forms/AllItems.aspx?RootFolder=%2Fgas%2Dcontrol%2FShared%20Documents%2FActivities%2FSmsMail%20notification&FolderCTID=0x012000852472212F969B4E9850475116730F2C

 

Archiving/Smoothing/Database

To study Gianluca's presentation in particular for what concerns deadband for driver and database

Is it possible to compare the archiving configuration of a running system with the configuration stored somewhere? for example before deploying an upgrade. Michal to check with Marc - this is done in reverse engineering tool - currently whatever is set in the production is consider as correct configuration - Geraldine would like to keep it that way to make sure in future we are not going to loose anything - so the idea is to fix the archiving problem directly on the production system withot changing the reverse engineering tool rules

can we have a warning for variables archiving too much? Michal, Gianluca to investigate

 

Gas System Overview (Exhaust++)

Check what is possible in terms of statistics in winccoa -> Michal, Gianluca with Geraldine and CERN winccoa trainer - winccoa is not meant to be data analytic tool - solution from BE-ICS to be explore as potencial candidate for data analytic for GAS and CO2 cystems

Can we have the sum of blocked alarms? -> Michal - I checked the code and we have it already 

 

AoB 

Examples discussed:

CMS:

-Why PGS is half green.

Michal to check, describe and write few lines of proposal (JIRA:GCSUR-111)

-DT analysis is half green due to normal cycle --> DT top module is half green --> no good representation

Ticket open -TICKET NUMBER-. Just to follow. Modif will be done once back (JIRA: GCSUR-112)

 

-Humidifier DT forced off. 

Introduce parameters to select if module is needed or not for run.

Change should not require stop/start of the system. Can we apply the same to purif? why today stop/start is required? (JIRA Purifier:vGCSUR-110) (JIRA humidifier: GCSUR-113)

Check if for RPC we have a parameter (for humidifier or for change without stop/start? I don't remember...) - Michal: there is no parameter,global ticket for humidifier modif already created (JIRA: GCSUR-113)

-Propagation of alarms to main pco requires creation of new alarms. True?

-Remove notification for propagation of alarms from child to parent -> Michal

-Alarm during maintenance: Michal to propose a solution to not stop interlocks but only notifications

How other groups are dealing with the same problem? -> Michal

ALICE:

ALIHMP_Di_61InPresAI 

ALIHMP: red = backup active --> modify to something else: full green backup active. Full red backup pressure error, empty green during run when backup is not used.

--> it will be included during TPC upgrade (test will be needed) (JIRA:ALGCS-729)

NA62:

Straw: Backup alarms alarms added on PT6x79 and 6x75. To add also text.  (JIRA:GNA-94)

RICH: Implementation warning/alarms on PT6125 ongoing -> Michal (JIRA:GNA-95)

CEDAR: Implementation warning/alarms on PT6104 or PT6105 to be defined -> Roberto

 

Description of alarms to be studied for the future. It requires new functionalities to be developed by SCADA team. Michal to add Roberto to jira ticket and to inform about future developments. is it correct? -MICHAL TO CHECK-(JIRA: ENS-26733 / UCPC-1052)

 

What is TIP? Michal to prepare a presentation for next meeting (JIRA:TIP-161)

 

Web page for main parameters to be developed by Gianluca. Now possible because all data are on GPN/DIP

 

Plots issues (some):

-auto-range (add details)

-problem saving plot config for some users (is it now solved?)

-changes are lost when zoom -> to be monitored (is it known/reported? -> Michal)

-in zoomed view for "dynamic trend" if I add something I have to close and open again -> to be monitored (is it known/reported? -> Michal) (JIRA: ENS-27012 )

 

Recipes not active info/warning.

As discussed in the past, something to prevent start-up of a system if recipes are not active

can be implemented before Run3? -> Michal to check -MICHAL TO CHECK- under discussion with Geraldine GCSSCD-202

 

Alarm/warning levels from winccoa

Is it possible to have a kind of download table --> from recipes? -> Gianluca, Michal --> to be discussed with Geraldine. News by mid-may (now busy with RICH2)

 

CANbus monitoring tools --> Gianluca, Michal (+ Marc?)

Canbus monitoring: Ethernet-CAN X model from anagate (Analytica). (up to 4 webmases can be connected)

http://www.anagate.de/en/products/AnaGateCANX4.htmhttp://www.anagate.de/en/products/AnaGateCANX4.htm

The integrate web interface of the CAN interface looks like the following. (you can see if there are errors on the bus, or lost packets, etc..) from set-up installed in BE-ICS lab (no data is taken at the moment) http://128.141.159.207/index.cgi?site=status

Geraldine will try to re-establish a setup in the lab with ELMB CANopen OPC UA such that we can see full picture

 

Find a way to have an effective communication with BE, other users and software development:

Michal reported that Lukasz is rapresenting "our" section but we never had discussion with him. Are there dedicated meeting? How often? Can we organize meeting to give feedback to Lukasz?...

15:20 → 15:40 Alarms from non-LHC experiments: summary

Speaker: Mara Corbetta (Universite Claude Bernard Lyon I (FR))

nonLHCalarm.png

nonLHC alarms
> received by: Roberto, Mara (students phone)
> about 2-4 alarms per week (current safe-mode status, no cloud)
> no difference between alarms - warning 

----------------------------------------------------------

link to general alarms summary 

comments: https://espace.cern.ch/gas-control/_layouts/15/WopiFrame.aspx?sourcedoc=/gas-control/Shared%20Documents/Activities/SmsMail%20notification/SMSAlarms-Actions.docx&action=default

stats: https://espace.cern.ch/gas-control/_layouts/15/WopiFrame.aspx?sourcedoc=/gas-control/Shared%20Documents/Activities/SmsMail%20notification/Summary_Alarms.xlsx&action=default

15:40 → 16:00 Alarms notification: egroup structure for LHC and non-LHC

Speaker: Michal Zbigniew Zimny (CERN)

16:00 → 16:20 Archiving/smoothing update

Speaker: Gianluca Rigoletti (Universite Claude Bernard Lyon I (FR))

GCS 4 Jun 2020.pdf

Link to summary presentation:

https://docs.google.com/presentation/d/1Cz0n9Yaqrs8eXkdx9dbMqU3lkTTiAD6iU_nLybcSuBI/edit?usp=sharing

16:20 → 16:40 Database usage

Speakers: Gianluca Rigoletti (Universite Claude Bernard Lyon I (FR)), Michal Zbigniew Zimny (CERN)

16:40 → 17:00 Exhaust++

17:00 → 18:00 aob

Speakers: Beatrice Mandelli (CERN), Gianluca Rigoletti (Universite Claude Bernard Lyon I (FR)), Mara Corbetta (Universite Claude Bernard Lyon I (FR)), Michal Zbigniew Zimny (CERN), Roberto Guida (CERN)

Mara 04/06

CMS RPC Gas Loss Alarm
> the loss can be negative: can we have alarms on ABS value? 
> not used channels > disable from reading? disabled channels are automatically excluded? 
> blocked alarms on leaking channels > is it correct to block from Alarm tab of distribution PCO?

------------------------------------------------------------------------------------------------------------

Gianluca:

• Data extraction from winccoa archiving. Not needed as it is the same data from Timber. Solved
• Fast method to check and set deadband, smoothing (winccoa and plc).
  • It's not fast but it's mostly automated, except the export of the config files. Presentation here: https://grigolet-eos.web.cern.ch/grigolet-eos/presentations/WinCCOA/DPE%20configs%20ALICE.slides.html
• Fast method to check alarms and warning levels
  • I could only find the data on CPC spec files but I couldn't find if there is the possibility to dump them from WinCCOA
• Improve the text in sms messages for alarm/warning to have a better understand of what is the issue
  • Michal?
• Alarms handling on elog:
  • I tested a script and dumped some alarms here: https://elog-epdt.web.cern.ch/Alarms/
  • System quite complicated because of cern-sso but we can test it in case
• Quick export of warning, alarms, forced items etc for each system
  • Discussed with Michal, waiting for reply
• Import of CPC device configs on production
  • Deadbands are in the spec file, so I suspect we cannot do it (without stopping everything) unless there is a dedicated feature. Asked to Michal, waiting for reply
• Development of the webmass software for read-out of gas temperatures
• Development of the webmass software for monitoring of the CANbus quality
• Development of an overview page in winccoa to show the Profibus and CANbus quality
• Any new test system (like the one of r134a recuperation)

 

Mara:

ALARMS

• Some alarms are not showed in the main panel, you need to open the system to see it
  > it is because of “Child to Parent” alarm in the module that is blocked
• Some alarms cannot be identified
  > slides Monday meetings with all the alarms to reconstruct together
• Block alarms notification (not the alarm itself) during maintenance
  > we could have a button to set the system in maintenance, with timer?

GENERAL

• Have an overview of each system with relevant parameters to see the situation at once
  > like exhaust page, maybe a button next to “Gas System”
     exhaust parameters, mixer flow/concentrations, pump status/opening, …

PLOTS

• Save a trend you made in “Dynamic Trend”, not everybody can do it
  > for Gianluca works in some systems, check if now it works or if it is a problem only of some systems (upgrade/not)
  > screenshot and put in elog (assign to Michal)
• Make modification in “Plot Configuration”, it can be opened but modifications are not applied for not-admin access
  > check if it can be made accessible to everyone
• Scale/Range problems with new software in “auto”
• Zoomed plot does not keep scale/time range, not updated with newly added trends

 

Mara (07/05)
> flow value in mixer lines has no archiving
   is it necessary to keep? useful to trace back consumption (low+high flow), does not make much sense without archiving (reading is the same as XMFC