SUM-like functionality after WLCG Monitoring Consolidation

513/R-068 (CERN)



Show room on map
How to provide the SUM-like functionality after the WLCG monitoring consolidation. The two alternatives that were being discussed: o) Offer that functionality in the standard SSB web interface o) Modify the current SUM to get the data from SSB (instead of SAM)
Present: Luca, Edward, Costin, Pablo, Stfan, Andrea, Ivan, Julia, Maarten, Marian, Markus, Alberto
Phone: David C., Laura
Apologies: Lionel, Pepe

Minutes taken by Pablo

- Review of the document (Pablo):

All the comments that had been submitted during the previous two weeks were combined in the version 1.0 of the document (attached to indico). The next step is to define the tasks in JIRA

For that, the responsible for a task should send the amount of time and dependencies of their tasks

-SUM functionality (Ivan)

Ivan went through the presentation and demo.

Ivan: Another difference between the current system and the new is that in the new we can select the metrics executed with different FQAN independently.

Ivan: At the moment, the availability is calculated as 'good/(good+bad)'
Andrea: Is this the same procedure that is applied in the SAM official reports?
Ivan: No, for that we still have to include the maintenance

Maarten: Why are there so many gaps in the plot of the tests?
Ivan: The validity that we have configured for the metrics is 3 hours (instead of the 24 hours that is currently used by SAM). The CE tests sometimes do spend more than that time in the queue

Andrea: Does the maintenance of two different UIs involve a lot of effort?
Ivan: Since Jarka left in March, there were no updates needed for the SUM UI
Maarten: So, if we maintain the interface without adding more enhancements, the amount of work is minimal.

Julia: SSB is already quite generic. Adapting it also to SUM might be too much
Andrea: I suppose that the merge of the backend is the biggest gain.
Julia; Exactly, the data is the same, independently of the UI used for the visualization. That's a a big improvement over the current situation.
Maarten: This strategy seems to be the best approach. We could revisit it in the future to see if later on it still makes sense

There is a consensus that it is better to keep two UI, instead of modifying the SSB UI to fit the SUM use case.

Markus: Which system(s) do we stop?
Julia; This means that we stop MyWLCG

-Discussion on the name of the project (following Alessandra's email)
SAM/SUM is too confusing for the end user.
Andrea suggested to use SAM also as the name of the new system, so that users will not be confused.
Pablo suggested to add a list the version number (SAM3), so that there is an easy way of specifying if we are referring to the original SAM, the current SAM or the new SAM.

Andrea: Any news on CondorG?
Marian: The CREAM-CE probe started working yesterday. For the CondorG, there has been some progress, and there are still things to do

-Next meeting:
It will be on the 10th of January, 14:00. Three topics:
  * Update on CondorG
  * Project plan on JIRA
  * REBUS functionality

Maarten: What do we have to discuss about REBUS?
Markus: The goal is to reduce the long-term work needed for maintaining all our applications.
Julia: Do we have to take care only of the visualization, or also the storage
Markus: Both.
Julia: Then, the use case is very similar to the SSB

There are minutes attached to this event. Show them.
    • 2:00 PM 2:20 PM
      Review of version 1.0 of the report 20m
      Speaker: Pablo Saiz (CERN)
    • 2:20 PM 2:40 PM
      SUM functionality in the WLCG-MON 20m
      Speaker: Ivan Antoniev Dzhunov (CERN)
    • 2:40 PM 2:55 PM
      Discussion 15m
    • 2:55 PM 3:00 PM
      Next meeting 5m