• VMs require 8 GB for parsing and processing the data from the dashboard. Probably the script can be improved, and if we need more performance then we could ask the dashboard team to pre-process the data or provide a custom API.
  • disambiguation of accesses: can that be provided at source, so we don't have to even account for it, and can just drop the packets?


  • xrootd: how to expand to world-wide access?
  • how do we caintain a set of GLED collectors? One per xrootd re-director?
    • GLED should not try to access directly the AMQ broker, should cache locally and have a consumer that read/sends-to-AMQ
  • How many MonALISA agents are deployed?
  • Does CMSSW popularity scale, if there's only one UDP collector?
  • Can we harmonise the UDP collection implementations? Long term...?
  • CMSSW Generic File Monitoring ( how do we monitor lost packets?

Action items

  • verify and validate information from CRAB3
  • find out how to submit a few verification jobs, or use Hammercloud jobs that test CRAB2 & CRAB3
  • refreshing the materialised view: can be done automatically?


  • evaluate how to install the services
  • evaluate the structure of the RPMs


  • vocms41 for Victor in production, CRAB2
  • vocms42 for Web Data Popularity and Victor
  • vocms43 same as vocms42
  • vocms44 for CMSSW popularity

Day two...

cron job ran overnight successfully on new VM, though as root.
  • maybe a page with status reports showing that it ran properly or not?
  • could run the cron (for CRAB) more frequently. CMSSE and xrootd are continuous, with latency of only a few minutes.


  • started reporting issues to github
  • hardcoded database scheme: use synonyms instead
  • need to be able to build development databases and deployments
  • list of issues in victor from Nicolo' now reported in github issues, started working on them
  • now have working spec files for victor monitoring and django, committed in DDM github until they're ready to go to cmsdist
  • Domenico will add tickets based on his issues
  • now have four voboxes and we have victor agents & crab popularity running on one, victor & popularity web running on two others, and CMSSW popularity agent on the fourth
    • DNS alias for two web applications, now exposed to the web redirector and running in production. Can switch off the dashboard machines when we're satisfied that this is OK
    • victor and crab popularity running production too, so can also switch off IT machines when ready
    • CMSSW popularity, waiting for the port to be opened in the firewall, then test to see if it's working (add to cms-dpmon-collector DNS alias)
    • aim to remove IT machines from aliases next Monday
  • need to work on putting the web interfaces behind cmsweb.
    • aim for testbed depoyment after the August release of cmsweb, with production release in September if all goes well.
    • upgrade Django to latest version
    • check about use of mod_wgsi
    • break-out highcharts and datatables into separate packages

-- TonyWildish - 16 Jun 2014

Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r3 - 2014-06-17 - TonyWildish
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2021 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback