Operations Bulletin 300913

From GridPP Wiki
Revision as of 10:20, 30 September 2013 by Jeremy coles (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Bulletin archive


Week commencing 23rd September 2013
Task Areas
General updates

Monday 30th September

  • There is a provisional agenda for the GDB next Wednesday.
  • Michel has indicated a desire for WLCG to review cloud progress - particularly around security issues like identity used to execute user payloads (identity change?), traceability requirements, and policy update (topics like root account usage). A F2F meeting on these topics is likely to take place in December/January.
  • We plan to decommission several VOs (see this ticket for the background. Does anybody have any knowledge of oxgrid.ox.ac.uk? Was there a ticket following up on this VO?
  • From today 3 ARC and 2 CREAM CEs will be declared as in production under Condor at RAL.
  • There was an EGI OMB meeting last Friday.
  • Problems with an ATLAS pilot factory and also FTS3 issues last week led to a low number of jobs across UK Tier-2s on several days.
  • Registration for the WLCG workshop in Copenhagen (November 11th-12th) is now open. Please let Jeremy know if you have a specific interest in attending (to ensure travel requests can be accommodated).
  • GOCDB v5 is scheduled for release on the morning of Wednesday 2nd October. This is a major release.
  • Following the HEPiX meeting in Bologna, a configuration management working group is being established, focussed primarily on puppet. The goal is not to replace the wider puppet community with something HEP specific for general puppet activity, but to find areas of collaboration for HEPiX specific problems. The initial motivation for setting up the WG were to share information between sites that are deploying puppet, and in particular for services where ongoing YAIM support is unclear. For more information see the wiki or contact Ben Jones or Yves Kemp.

Tuesday 17th September

  • The EGI Technical Forum is taking place this week. Take a look at the agenda. Not all talks are yet online. There are training materials/sessions on security and Glue2 publishing validation.
  • The GPGPU working group in EGI is seeking more participants.
  • The SHA-2 deadline has moved to 1st December.
  • There is a new CHEP 2013 bulletin.
  • WLCG continues to run a biweekly operations meeting on Monday's and Thursday's. See the minutes here.
  • The final WLCG T2 reliability/availability report for August has now been released.
WLCG Operations Coordination - Agendas

Tuesday 17th September

  • The agenda of the next WLCG operations meeting is available here. The details of the agenda are not yet final. The participation of the Tier-1 contacts is being strongly encouraged, but also Tier-2 sites are welcome to listen in and contribute (via Vidyo).

Tuesday 2nd September

  • Middleware
    • New BDII release in the latest EMI-2/3 update, including better GLUE-2 support and security fixes. Sites should update all their BDII instances
    • New CVMFS version released for a security fix. Sites should upgrade or at least apply the hot fix in the above twiki
    • perfSONAR: sites should upgrade to the latest version, fixing many deployment problems
    • The end of support for dCache 1.9.12 has been postponed to September 30 due to a delay in releasing the SHA-2 compliant version in the dCache 2.2 series.
    • Consult https://twiki.cern.ch/twiki/bin/view/LCG/WLCGBaselineVersions
  • SHA-2
    • Discussion mostly dedicated to the experiments testing status. Atlas and LHCb have tested the services but not job submission yet. All experiments have been encouraged to test this.
  • SL6
    • T2 Done: 49/129 (Alice 11/39, Atlas 28/89, CMS 22/65, LHCb 13/45) -> 80/129 still to be done.
    • HS06: Reminder that sites are requested to run HS06 benchmark and update the value in the BDII. Increased values might be discussed at the WLCG MB.
    • EMI-3: voms-clients have been fixed and the latest version is in the PT repository but not in EMI-3 yet. Both CMS and Atlas work on DPM/dcache sites with this patch. (QMUL might want to give an update on Storm when they upgrade)
    • UK status: Liverpool to be finished soon, Bham in downtime to upgrade this week, Bristol and Sussex should be done by the 15/9/2013, RALPP 20/09/2013 and QMUL, Lancaster, UCL 30/09/2013
  • glexec
    • 55 sites still to respond they have attached the installation to SL6 upgrade.

Monday 12th August

  • There have been no recent meetings. The next is on 29th August.
      • Sites monitoring requirements: SUM tests not representing the real experiment status for example.
Tier-1 - Status Page

Tuesday 17th September

  • There have been some problems with the Top BDIIs. Following this one (of the three) was upgraded to the latest version of the software. However, a configuration problem in this node following this upgrade caused an additional problem on Friday.
  • The Condor batch farm (with SL6 WNs, both ARC & CREAM CEs) has had more nodes added to extend the testing. In parallel we have planned the upgrade of WNs to SL6 within the 'old' (Torque/Maui) batch farm.
Storage & Data Management - Agendas/Minutes

Monday 30th September

  • A DPM workshop is being organised in Edinburgh for 13th December. GridPP PMB anticipated covering travel for of order 10 UK sysadmins for this event. Interest should be indicated during the storage group meeting.

Tuesday 17th September

  • Perhaps someone could summarise the "Dark Data identification tools" thread on TB-Support?



Accounting - UK Grid Metrics HEPSPEC06 Atlas Dashboard HS06

Tuesday 13th August

Tuesday 23rd July

  • Sites moving to SL6 are reminded of the need to re-benchmark their WNs. Some sites have updated the wiki already and provide an idea of the performance change.
  • There is an ongoing PMB discussion about the timeline for the next Tier-2 hardware tranche. Please let Pete or Jeremy know if your site will benefit from a spend this financial year.

Tuesday 30th April

  • A discussion is starting about how to account/reward disk that is reallocated to LHCb. By way of background, LHCb is changing its computing model to use more of Tier-2 sites. They plan to start with a small number of big/good T2 sites in the first instance, and commission them as T2-Ds with disk. Ideally such sites will provide >300TB but for now may allocate 100TB and build it up over time. Andrew McNab is coordinating the activity for LHCb. (Note the PMB is already aware that funding was not previously allocated for LHCb disk at T2s).

Tuesday 12th March

  • APEL publishing stopped for Lancaster, QMUL and ECDF

Tuesday 12th February

  • SL HS06 page shows some odd ratios. Steve says he now takes "HS06 cpu numbers direct from ATLAS" and his page does get stuck every now and then.
  • An update of the metrics page has been requested.
Documentation - KeyDocs

See the worst KeyDocs list for documents needing review now and the names of the responsible people.

Tuesday 1st October

  • The approved VOs page has been updated with the newest data from the operations portal. Note that the VOMS records for LondonGrid now contain some alternative voms servers. The migration plan for use of these backup servers is now document here.

Tuesday 17th September

  • Little obvious change in the status table since last week.

Tuesday 3 September 2013

  • Proposal for "Instant UI", with the aim to produce a suite of documentation and software that will enable a new user to set up a UI and join the grid with the minimum of hassle. Doc will show for to admin a UI that can be used to submit jobs and retrieve output for a given set of users belonging to a given set of VOs. "Instant UI" is currently in consulation phase with GridPP admin community.
Interoperation - EGI ops agendas

Monday 30th September

  • There was an EGI ops meeting on 23rd September. See the agenda for more details.

Monday 16th September

  • The next meeting takes place on 23rd September at 13:00 (UK time).
  • UMD 3.2.0 was released last week. See the release page for more information.

Monday 2nd September

  • Yesterday's agenda. Attended by David and Raul.
gLite support calendar.


Monitoring - Links MyWLCG

Monday 30th September

  • David summarised the UK site's position on Nagios in an email last week as:
  • There is a desire for a monitoring solution that gave automatic notifications and links to further information, and didn't require additional webpages (which describes Nagios). We noted that Nagios could be used to import central nagios tests and repurposing them for local testing.
  • In addition, it would be useful if the further details could include details of the testing execution commands (even including the test itself) for local diagnosis.
  • We wondered whether (and where) there might be common ground with the WLCG Nagios project - while this may have been discussed, it would be useful to clarify this.
  • It's important to have a clear and documented messaging/transport layer for any solution that's decided on, for integration with future monitoring solutions.

Tuesday 23rd July

Tuesday 18th June

  • David C is taking feedback on the Graphite implementation presented at the HEPSYSMAN meeting. Also considering integrating Site Nagios.
  • Glasgow dashboard now packaged and can be downloaded here.
On-duty - Dashboard ROD rota

Monday 30th September

  • Quite a good week - with problems at only a couple of sites. No UK-wide problems.
  • Point to note for this week: several of the SHA-2 tickets are currently set to expire at the start of October, due to the original deadline.

Tuesday 17th September

  • Daniela is on duty this week.
  • A lot of alarms last week because of top bdii issue at RAL Tier1. lcg-* command uses topbdii to get se endpoints so this issue is affecting actual jobs as well.
  • SHA2 tickets still open for three sites.
  • SE issue at Durham is still going on. Open ticket.
  • Opened ticket for Sussex Apel issue. No response from the site.
Rollout Status WLCG Baseline

Tuesday 17th September

  • Chris sent in a report for Storm.

Tuesday 9th July

  • New EMI2 and EMI3 release yesterday. No staged rollout requests yet. Imperial upgraded their WMS and they have been somewhat shaky ever since.

Tuesday 18th June

  • New EMI3 CE coming into SR. Liverpool will test.
  • A lot of EMI3 testing done at Brunel.
  • EMI-3 testing page contains all issues I am aware off. It's a Wiki though, so if you find an issue, please put it in the appropriate category.

Tuesday 14th May

  • A reminder. Please could sites fill out the EMI-3 testing contributions page. This is for all testing not just SR sites as we want to know which sites have experience with each component.

References


Security - Incident Procedure Policies Rota

Tuesday 17th September

  • More information on the EGI/PRACE/EUDAT Joint Security Training event mentioned last week is now available.

Tuesday 3rd September

  • Security contacts and system staff of the partners of EGI, EUDAT and PRACE are invited for a joint security event from Monday, October 7 - Wednesday, October 9 in Linköping, Sweden. Monday and Tuesday will be a training event which should be of interest for all staff managing the systems which are part of our infrastructures. The second part of the meeting will consist of a discussion of security policies and the collaboration among the different infrastructures.
  • Rob H has moved to the Tier-1. Looking at options for security team.
  • At the next team meeting we should review our approach to 100percentit (just moved to uncertified).

Tuesday 20th August

  • Several sites showing up in pakiti this week.
  • Update on workaround discussed last week.


Services - PerfSonar dashboard | GridPP VOMS

Tuesday 1st October

  • PerfSONAR latency hosts configured to use the WLCG meshes should now have a traceroute measurement achive (MA) accessible from the GUI under 'Service Graphs' --> 'Traceroute'. Here is an example.

Tuesday 17th September

  • Upgrading/re-installing hosts to v3.3.1/mesh is only making slow progress.
  • There is a new view of the status between sites.
  • An outage at Manchester due to central switch maintenance means that VOMS is not going to be contactable for a period this morning. It is clear that we need the backup VOMS instances fully available to VOs - please can someone take a lead?
Tickets

Monday 16th September 15.00 BST</br> It's all hands on deck here at Lancaster tomorrow as we move the shared cluster to SL6 (and SGE). So I may or may not be in the meeting tomorrow, depending on how things go! I would have skipped the review this week, but we're not getting one next week due to GridPP so I'll squeeze in a quick look at the tickets.

So here are the tickets that caught my eye this week:

MOST NEGLECTED TICKET:</br> SUSSEX https://ggus.eu/ws/ticket_info.php?ticket=97139</br> Hiding at the back of the cupboard with the marmite is this ticket, assigned a week ago. Being a ROD ticket this is at risk of escalating, and the ROD can't really do anything unless the ticket is at least acknowledged. Assigned (9/9)

HYPERK</br> @RAL</br> https://ggus.eu/ws/ticket_info.php?ticket=96233</br> https://ggus.eu/ws/ticket_info.php?ticket=96235</br> Catalin reports that the additional voms servers have been enabled on the WMS and LFC at RAL, both tickets await some testing.</br> @GLASGOW</br> https://ggus.eu/ws/ticket_info.php?ticket=96234 Chris would have it known that HyperK are now in the Operations portal, but there's no rush in enabling them at Glasgow though. I don't know if there's a general call for the sites to support the new VO?

RAL</br> https://ggus.eu/ws/ticket_info.php?ticket=95996 (22/7)</br> This is the SHA-2 ticket, it's set In Progress but looks like it's waiting on work that is still forthcoming, so probably could be On Holded, or at least updated. In Progress (5/9)

https://ggus.eu/ws/ticket_info.php?ticket=97168 (9/9)</br> Duncan spotted problems with the londongrid lfc. The ticket was acknowledged, but no news since. In Progress (9/9)

GLEXEC TICKETS</br> We're approaching the period where a lot of the sites with open GLEXEC tickets said they'd be starting to roll out GLEXEC. I'll save looking at these till the October ticket review.

UNRESPONSIVE VO TICKETS</br> Should the remaining two of these be closed and replaced with fresh, VO decommissioning tickets?

Tools - MyEGI Nagios

Monday 30th September

  • Ewan has put together a slightly modified WLCG VO box, but the effect is of a UI that takes gsi ssh logins from people in one particular VO, but then can be used as a UI for other VOs once you're logged in. The idea is that anyone who would need access to a central UI machine (so, mostly not people in PP depts.) would join a special-purpose VO. See Ewan's TB-SUPPORT email on 23rd September for more details.

Monday 2nd September

  • Intermittent Nagios errors -> Imperial WMS and all the jobs going through it were failing with ‘no compatible error’. Some reports of ongoing issues. What is the direct impact?
  • MyEGI and gstat were also down last week.
  • Jens is testing SHA-2 compliance of components. The version of gridsite on the GridPP website is not compliant but SHA-2 will be supported with a move to a new server (when?).
VOs - GridPP VOMS VO IDs Approved VO table

Monday 2nd September

  • The next quarterly Tier 1 allocation/resourcing meeting is scheduled for Wednesday 18th September (after the weekly T1 meeting) the hardware requirements and fair-shares for the period October-December 2013 will be reviewed. It looks ahead over the next 12 month timeframe. Can all experiments/projects please let Pete G have any updates or requests to these numbers by Friday 13th September please?

Monday 19 August

  • EPIC
    • Support requested at Tier-1
    • Any other sites prepared to support them?
  • Catalogue synchronisation - Biomed working on it.


Monday 12 August

  • HyperK.org
    • VOMS servers set up (Manchester, Oxford, Imperial)
    • VOID card - stalled on a homepage.
    • WMS set up (Imperial) - awaiting Glasgow, Ral
    • Site set up (QMUL)
    • LFC - in progress
    • CVMFS - considering
  • SNO+
    • Dirac set up for some CEs
  • Epic
    • Doing stuff
  • ngs.ac.uk VO - any reason to keep it?
  • Software areas for SL6
    • Are we keeping the same areas as sl5?
    • What about the software tags?
    • Push CVMFS?
Site Updates

Actions


Meeting Summaries
Project Management Board - MembersMinutes Quarterly Reports

Empty

GridPP ops meeting - Agendas Actions Core Tasks

Empty


RAL Tier-1 Experiment Liaison Meeting (Wednesday 13:30) Agenda EVO meeting

Wednesday 18th September

  • Operations report
  • During this last week the '08 and '09 batches of worker nodes were successfully added to the Condor farm which has run OK since. We are finalising plans regarding upgrading the remaining WNs to SL6.
  • Top-BDII issues during last Wednesday (11th Sep) and again on Friday (13th Sep): The (then) current version of the Top_BDII software has performance issues and can become unresponsive. These can sometimes (but not always) be alleviated by restarting the service. On Thursday (12th Sep) one of the Top-BDII nodes (lcgbdii04) was upgraded to a newer version of the software. Although this new version is responsive, it introduced a new bug causing it to not contain information about various Site-BDIIs. It was this bug that affected our Top-BDII service on Friday - which was fixed by removing the particular node from the alias.
  • There was a problem in the Oracle database behind Castor that called out last night (17/18 Sep). This is a problem of which we were aware and the fix is to change an Oracle parameter. A database restart was needed for this and Castor has been stopped (and batch jobs paused) for this to take place. (Note: This intervention was completed successfully and all services restored shortly after the meeting ended.)
  • RAL will move to SuperJanet 6 on Tuesday 1st October.
WLCG Grid Deployment Board - Agendas MB agendas

Empty



NGI UK - Homepage CA

Empty

Events

Empty

UK ATLAS - Shifter view News & Links

Empty

UK CMS

Empty

UK LHCb

Empty

UK OTHER
  • N/A
To note

  • N/A