Hi,

The first algorithm will be replaced by the VO-based (second) algorithm starting this month (Jan, 2014). So we have been working with WLCG to identify what them measure. Grid jobs for the new algorithm require outbound TCP connection from batch nodes which we didn't have until Oct/Nov. It takes many iterations to identify these types of issues. I am in close contact with those people at WLCG to straight out individual benchmarks but so far we can't identify exactly which benchmarks are used in the reports (for WT2). So getting the month report in the new algorithm timely is important to us.

regards,
Wei Yang  |  [log in to unmask]  |  1-650-926-3338




On Jan 9, 2014, at 9:32 AM, "Young, Charles C." <[log in to unmask]> wrote:

> Hi Wei, 
> 
> Happy new year. Thanks for the report. I have some questions. The file https://espace.cern.ch/WLCG-document-repository/ReliabilityAvailability/Tier-2/2013/WLCG_Tier2_OPS_Nov2013.pdf and the "new VO-based reports" can paint very different pictures. I am looking at availability history (starting at page 3 and page 5 respectively). Scroll down to US-WT2. The numbers in the first report are: 94%, 94%, 1%, 93%, 97% and 97%, while the numbers in the second report are: 84%, 84%, 18%, 1%, 0%, 0%. I don't believe we were totally unavailable for August, September and October — the last three numbers. The second set must be screwed up. 
> 
> We are not the only one with massive differences. On the other hand, there are also sites with relatively small differences between the two reports. I can see that they differ in the algorithms used to compute availability. They must also have a different calculation for availability because the second numbers don't agree. I would not argue that the first report's algorithms are perfect, but I would certainly argue that the second one's are no good. 
> 
> This may be well known and understood. It's just that I haven't looked carefully for some time. Cheers.
> 
> Charlie
> --
> Charles C. Young
> M.S. 43, Stanford Linear Accelerator Center      
> P.O. Box 20450                                        
> Stanford, CA 94309                                      
> [log in to unmask]                                
> voice  (650) 926 2669                        
> fax    (650) 926 2923                      
> CERN GSM +41 76 487 2069
> 
> From: <Yang>, Wei Yang <[log in to unmask]>
> Date: Thursday, January 9, 2014 12:04 AM
> To: atlas-sccs-planning-l <[log in to unmask]>
> Subject: Fwd: T2 Reliability & Availability - November 2013
> 
> Begin forwarded message:
> 
>> From: WLCG Office <[log in to unmask]>
>> Subject: RE: T2 Reliability & Availability - November 2013
>> Date: January 8, 2014 at 11:49:29 PM PST
>> To: "project-wlcg-cb (Members of the WLCG CB)" <[log in to unmask]>
>> Cc: "project-lcg-gdb (LCG - Grid Deployment Board)" <[log in to unmask]>, "sam-support (SAM support)" <[log in to unmask]>, "[log in to unmask]" <[log in to unmask]>, "[log in to unmask]" <[log in to unmask]>
>> 
>> Dear all,
>> 
>> The final T2 reliability & availability report for November is here:
>> 
>> https://espace.cern.ch/WLCG-document-repository/ReliabilityAvailability/Tier-2/2013/WLCG_Tier2_OPS_Nov2013.pdf
>> 
>> Please also check the new VO-based reports for November here, which include all re-computation requests received.
>> 
>> Kind regards,
>> Cath
>> 
>> -----------------------------------------------
>> WLCG Office
>> IT Dept - CERN
>> CH-1211 Genève, Switzerland
>> www.cern.ch/wlcg
>> From: WLCG Office
>> Sent: 05 December 2013 09:28
>> To: project-wlcg-cb (Members of the WLCG CB)
>> Cc: project-lcg-gdb (LCG - Grid Deployment Board); sam-support (SAM support); [log in to unmask]; [log in to unmask]
>> Subject: T2 Reliability & Availability - November 2013
>> 
>> Dear all,
>> 
>> The draft T2 reliability & availability report for November is now available at:
>> 
>> http://sam-reports.web.cern.ch/sam-reports/2013/201311/wlcg/WLCG_Tier2_OPS_Nov2013.pdf
>> 
>> Please verify your data and send comments to WLCG Office by 15 December.
>> 
>> All requests for recomputation must be sent via GGUS within 10 calendar days of this e-mail being sent - further details here.
>> 
>> Reminder: please check the new VO-based Availability Reports - November reports are here.
>> 
>> Kind regards,
>> Cath
>> 
>> 
>> -----------------------------------------------
>> WLCG Office
>> IT Dept - CERN
>> CH-1211 Genève, Switzerland
>> www.cern.ch/wlcg
> 
> 
> Use REPLY-ALL to reply to list
> To unsubscribe from the ATLAS-SCCS-PLANNING-L list, click the following link:
> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=ATLAS-SCCS-PLANNING-L&A=1

########################################################################
Use REPLY-ALL to reply to list

To unsubscribe from the ATLAS-SCCS-PLANNING-L list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=ATLAS-SCCS-PLANNING-L&A=1