MinutesSep27

Introduction

• Attendees: Charles Bacon, Anand, Jeff, Suchandra, Rob, Burt, RobQ, Tom, Neha, Steven Timm, Gabriele, John Hover
• Apologies: none
• Coordinates: Thursdays, 2:30pm Central; 510-665-5437, #1212
• Previous meetings, MeetingMinutes

ITB 0.7.0 background

• ReleaseDocumentation
• SiteValidationTable <--- Please update w/ latest

VDT and ITB status

• VDT 1.8.1 released
• ITB 0.7.1 is ready to be released
• Site to update: UC, FNAL, CIT-LIGO, BNL, LBL

WS-GRAM testing (Charles Bacon)

• See updates to the validation table - expanded

OSG Client Testng (Charles)

• See problem with voms file

VO validation status (Charles, Abhishek)

Here is an update for this week -

As you know, ATLAS completed validation on UC_ITB. From CMS, I am requesting Haifeng Pi to see if he can possibly validate CMS jobs/storage on UCSDT2-ITB1. nanoHUB and LIGO performed another round of validation, and had more results this week. Charles is in contact with Steve/nanoHUB and 2 sites that did not succeed for nanoHUB jobs.

This time, we are trying not to bother VOs too much (keeping in mind the email discussion we had a few weeks ago, and advice from Ruth and Frank). I will send a short reminder to CompBioGrid?, Dzero, DOSAR, Engage, STAR tomorrow. Btw, did you hear back from Mike for iVDGL?

regards, Abhishek

• Need to check CDF contact
• Horst will post some info for DOSAR on the page
• iVDGL - keep for now, since we have more time
• LIGO - still failing on three sites. UC_ITB - failing, but not sure why.
• Nanohub - auth problems on BNL
• SDSS - finished all tests
• STAR

• From Charles

ATLAS:  Abhishek informed me atlas run on the local UC_ITB site, but have not seen post result to the twiki page yet
CDF: not hear anything back yet
CMS: Abhishek is going to inform his colleague to run a simply job on local CMS site at UCSDT2-ITB1
CompBioGrid: not hear anything back yet
DOSAR: not hear anything back yet  (I am not sure whether Horst belongs to DOSAR, and whether he can have some influence on the testing)
Dzero: not hear anything back yet
Engage: not hear anything back yet
FERMILAB: finished all test
IVDGL: not hear anything back yet
LIGO: validate both CE and SE, all SE site test successful, currently 3 site fails on the CE test (UC_ITB, CIT_ITB_1, UCSDT2-ITB1)
nanoHUB: only validate CE, two sites have problems (BNL_ITB_Test1, UCSDT2-ITB1)
SDSS/DES: finished all test
STAR: finished all test, will validate on the fail sites again soon. Charles have sent the fail messages from STAR to all fail site admin for attenttion. 

VORS validation status (RobQ)

• Have updated validated page. Still will address point about access points.

LCG interoperability status (Burt)

• No issues now relating to BDII
• Tony's script that checks the web display from CEMon versus what comes from BDII
• Could we use the lcg-sites-info?
• Once BDII was fixed, moved forward.
• Andreas may be installing worker node tarballs.
• Site validation table up to date
  • 6 sites succeeeding - hellow world and srmcp
  • IUPUI - moving
  • UCSD - power outtage, gip patch applied?
  • LBL - site down
  • UC_ITB: configure gip advertises all the PBS queues advertised. Not hackable... gums-host-cron resets this.

BDII Aggregator (Anand, Tim)

• Was publishing stale information, affecting interoperability
• BDII-update process was not running
• Now working correctly - creating a document for Facilities group
• Also working on an monitor for three processes
• A couple of GIP related updates

ReSS validation status (Gabriele)

• Rob, Alain, Gabriele, Suchandra met to discuss how to make sure ReSS information is accurate
• Mechanism to check specific class ads - eg., number of nodes .ne. zero.
• Defined centrally, with static attributes; evaluate with a condor status command to evaluate expressions
• Basic ideas have been shown to work.

RSV validation status (Arvind)

• Happy with version in VDT 1.8.1

Documentation (Anne)

• Documentation hub being reviewed and table re-organized
• VDT team will begin to help starting next week

Site Status Reports (Site Admins)

• BNL - will update to ITB 0.7.1
• FNAL - have already tried VTB based on 1.8.1; will check carefully RSV.
• UC - going to be updating to 0.7.1 - will be some downtime next week
• CIT ITB - no report
• CIT LIGO - problem with RSV probes; re-installing VDT 1.8.1/ITB 0.7.1
• TTU - no report
• UCSD - no report
• Purdue - no report

Squid validation status

• Carryover from last meeting: OU will volunteer as a site for squid, will report next week

Update as of 9/26/07: Squid validation exercise has been completed successfully from the ITB (OUHEP_ITB) server and the WNs. Squid was installed separately from the OSG-0.7.0 main installation and the server started. configure-osg.sh was ran and the details of squid provided. configure_squid failed complaining that the squid installation wasn't found under the main osg-0.7.0 installation. But strangely this didn't seem to affect squid in any way. Running configure-osg.sh updated the osg-attributes.conf file with the OSG_SQUID_LOCATION attribute. wget test was done on the server and it succeeded (MISS followed by a HIT). wget tests were also done by submitting jobs to the ITB cluster through globus-job-run (so that the job executes on a WN) and it did pass as expected (OSG_SQUID_LOCATION was understood and MISS was followed by a HIT). Feedback has been provided to Dave Dykstra about these tests and it has been concurred that the validation was successful. A GOC ticket has been opened (ticket number 4150?) for the configure-osg.sh issue as discussed above. Thanks to Dave, Steve & Horst for the help during the tests. -Karthik

• Problems when squid is installed outside of the OSG installation
• Suchandra will address these before releasing ITB 0.7.1.

AOB

• Next week we will have an OSG Sites Coordination meeting in this timeslot (no ITB meeting)