sPHENIX tracking discussion

[Anthony Frawley <afrawley AT fsu.edu>]

Hi Chris,

If I remember correctly, the choice of which nodes are written out is made in a macro. Can we just look at the macro to see?

Tony

---

From: sPHENIX-tracking-l <sphenix-tracking-l-bounces AT lists.bnl.gov> on behalf of Christof Roland via sPHENIX-tracking-l <sphenix-tracking-l AT lists.bnl.gov>
Sent: Tuesday, March 22, 2022 12:31 PM
To: pinkenburg <pinkenburg AT bnl.gov>
Cc: sphenix-tracking-l AT lists.bnl.gov <sphenix-tracking-l AT lists.bnl.gov>
Subject: Re: [Sphenix-tracking-l] MDC2 Time

 

Hi Chris, 

agreed.

do we have a way of checking how many of which type of object go into the 3+GB track file?

Cheers

   Christof 

On 22. Mar 2022, at 16:07, pinkenburg <pinkenburg AT bnl.gov> wrote:

Hi Christof,

we need to subtract the copying, no argument there. But this does not scale per event, it is an offset to copy a 2-3GB file to the local disk and copying a 3+GB file to lustre (which then begs the question of the content of the tracks files - this just doesn't look right, even though the TRKR_HITSET node is not saved).

Chris

On 3/22/2022 6:09 AM, Christof Roland wrote:

Hi Chris, 

we should look into this soon. If we spend 10sec processing the event and 6sec copying we have a problem for the computing review.

Cheers

   Christof 

On 21. Mar 2022, at 23:04, pinkenburg via sPHENIX-tracking-l <sphenix-tracking-l AT lists.bnl.gov> wrote:

Hi Joe,

the times I report come straight from the condor logs, jobend - jobstart. Things like file transfers or startup are included in this (and these are a few GB files which take a while). I need to split this up a bit to measure this properly, but I can see that the copying can take time

Chris
 

On 3/21/2022 4:38 PM, Osborn, Joe via sPHENIX-tracking-l wrote:

Hi all,

 

I got sidetracked by some other things this afternoon, but here are some slides showing the timing breakdown of the 16s that Chris reported this morning. Interestingly, the scripts I have that mine the log files for the individual module times show a better total than 16s per event, so either I’m screwing something up (always possible) or there is something that is taking longer elsewhere that we need to understand.

 

Note that to produce these I took the lines from the logfiles that say: 
*_TOP: per event time (ms)

 

Joe

 

---------------------------

 

Joe Osborn, Ph.D.

Associate Research Scientist

Oak Ridge National Laboratory

osbornjd AT ornl.gov

(859)-433-8738

 

_______________________________________________
sPHENIX-tracking-l mailing list
sPHENIX-tracking-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-tracking-l

-- 
*************************************************************

Christopher H. Pinkenburg	;    pinkenburg AT bnl.gov
				;    http://www.phenix.bnl.gov/~pinkenbu

Brookhaven National Laboratory	;    phone: (631) 344-5692
Physics Department Bldg 510 C	;    fax:   (631) 344-3253
Upton, NY 11973-5000

*************************************************************

_______________________________________________
sPHENIX-tracking-l mailing list
sPHENIX-tracking-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-tracking-l

-- 
*************************************************************

Christopher H. Pinkenburg	;    pinkenburg AT bnl.gov
				;    http://www.phenix.bnl.gov/~pinkenbu

Brookhaven National Laboratory	;    phone: (631) 344-5692
Physics Department Bldg 510 C	;    fax:   (631) 344-3253
Upton, NY 11973-5000

*************************************************************