Tech Meeting Minutes 20101104

From ScotGrid

Present: Graeme, Sam, Mark, David, Stuart, Mike, Peter

http://indico.cern.ch/conferenceDisplay.py?confId=112740

Backlink: http://www.scotgrid.ac.uk/wiki/index.php/ScotGrid_Technical_Meetings

Table of contents
1 Security
2 Site Issues
3 Gstat
4 Site Updates
5 AOB
[edit]

Security

  • Latest kernels fix all but the second RDS module vulnerability, but RDS is disabled on all sites.
  • Glasgow
    • All systems patched.
  • ECDF
    • We think patched
  • Durham
    • Will patch this weekend. ACTION.
[edit]

Site Issues

  • Glasgow
    • Substation problems caused two power cuts on Friday. Cluster was back up and running in the evening, but put into downtime until Monday.
    • Knock-on effects on viglen disk servers (many PSUs died). Sam managed to get all the servers back running early in the week.
    • Open ticket about WMS not renewing proxies, https://gus.fzk.de/ws/ticket_info.php?ticket=63656. David investigating.
[edit]

Gstat

  • Reviewed http://gstat-wlcg.cern.ch/apps/capacities/sites/. Glasgow still has a wrong total (15k should be 21k). This is entered manually into the information system (via YAIM, usually). ACTION: Mark to correct this and also publish per-VO shares at Glasgow (only for LHC + others).
  • Peter asked where he could get info on site pledges. ACTION: Graeme to forward GridPP spreadsheets which give this. G will also forward WLCG MoU document.
[edit]

Site Updates

CREAM:

  • Glasgow will review the plan for a second CREAM CE, but will likely go ahead.
  • Durham - keen to try CREAM but installation likely to be next year.


glite-APEL:

  • Glasgow: next week.
  • ECDF: TBD.
  • Durham: After Glasgow and Edinburgh. Peter said they should be able to add a new VM to do the transition in a similar way to Gla & Ed.

VMs

  • Durham not very happy with VM setup - unstable. Will discuss this topic at F2F meeting in January.
[edit]

AOB

Retrieved from "http://www.scotgrid.ac.uk/wiki/index.php/Tech_Meeting_Minutes_20101104"