GSI Forum
GSI Helmholtzzentrum für Schwerionenforschung

Home » CBM » CBMROOT » General » Mass production DEC08 with full ECAL
Mass production DEC08 with full ECAL [message #7618] Wed, 10 December 2008 20:19 Go to next message
Volker Friese is currently offline  Volker Friese
Messages: 365
Registered: April 2004
Location: GSI CBM
first-grade participant
From: *gsi.de
I. Korolko requested a mass production including the full ECAL. This will be done with the release DEC08 (once it is there). It will be the first use case of CBM-GRID. This forum topic is intended to discuss issues related to this production.

Requested are:

  • Minimum 100 k events. This would require about 200 CPU days and about 0.5 TB storage capacity.
  • Maximum 1 M events, requiring 8 CPU years and 5 TB storage.


Production should start before christmas. The participating sites are GSI, Dubna, and potentially ITEP.

Open questions are:
  • What is the geometry to be used?
  • Which reconstruction steps are required?
  • What is the status of CBM-GRID at the participating sites?

Re: Mass production DEC08 with full ECAL [message #7635 is a reply to message #7618] Tue, 16 December 2008 13:39 Go to previous messageGo to next message
Mikhail Prokudin is currently offline  Mikhail Prokudin
Messages: 5
Registered: March 2006
Location: Moscow, ITEP
occasional visitor
From: 62.63.90*
Hello Volker,

ecal_FullMC.geo should be used.

All steps of reconstruction should be done. (I mean clustering, photon reconstruction and matching).

Reconstruction macro [message #7636 is a reply to message #7635] Tue, 16 December 2008 15:37 Go to previous messageGo to next message
Volker Friese is currently offline  Volker Friese
Messages: 365
Registered: April 2004
Location: GSI CBM
first-grade participant
From: *gsi.de
Dear Misha,

the simulation part is clear, then. Can you provide a reconstruction macro with the ECAL reco tasks?
Transport geometry [message #7637 is a reply to message #7635] Tue, 16 December 2008 15:42 Go to previous messageGo to next message
Volker Friese is currently offline  Volker Friese
Messages: 365
Registered: April 2004
Location: GSI CBM
first-grade participant
From: *gsi.de
Dear Misha,

I assume you want the standard geometry for electron measurement. The only question then is whether MVD should be included (as for low intensity running) or not (as in the case of maximal interaction rate).
Re: Transport geometry [message #7638 is a reply to message #7637] Tue, 16 December 2008 15:58 Go to previous messageGo to next message
Mikhail Prokudin is currently offline  Mikhail Prokudin
Messages: 5
Registered: March 2006
Location: Moscow, ITEP
occasional visitor
From: 62.63.90*
Dear, Volker,

I have attached the script to this letter. Please, verify the part preceding the ecal.

The problem is, that I need a new subdirectory ecal in parameters directory to put there configuration files for reconstruction. I also need to put some where a shower library and S-curve library (~200Mb, not in SVN for sure), which is necessary for reconstruction process. (I have written a mail to you, but it was ignored --- as usual:-)
  • Attachment: run_reco.C
    (Size: 19.14KB, Downloaded 393 times)
Re: Transport geometry [message #7639 is a reply to message #7637] Tue, 16 December 2008 16:00 Go to previous messageGo to next message
Mikhail Prokudin is currently offline  Mikhail Prokudin
Messages: 5
Registered: March 2006
Location: Moscow, ITEP
occasional visitor
From: 62.63.90*
Dear, Volker.

I don't know the answer for this question. We should ask Ivan.
ECAL parameter subdirectory [message #7640 is a reply to message #7638] Tue, 16 December 2008 16:48 Go to previous messageGo to next message
Volker Friese is currently offline  Volker Friese
Messages: 365
Registered: April 2004
Location: GSI CBM
first-grade participant
From: *gsi.de
Dear Misha,

not ignored - I have put you in my spam filter.... Razz

I have created the directory parameters/ecal in trunk (permissions will be set asap). I suggest to put the large files in the input directory. Please provide a mean to download it - e.g. sftp it to somewhere at GSI.
Test with full ECAL on desktop [message #7641 is a reply to message #7618] Tue, 16 December 2008 16:52 Go to previous messageGo to next message
Volker Friese is currently offline  Volker Friese
Messages: 365
Registered: April 2004
Location: GSI CBM
first-grade participant
From: *gsi.de
I have run the simulation with full ECAL on my desktop (sarge32) without problems (5 test events only). No crash as in the case of fastEcal occured Smile

Consumption:
2,903 CPU s (about 600 CPU s per event)
28 MB output file (about 5 MB/event)

Re: Test with full ECAL on desktop [message #7642 is a reply to message #7641] Tue, 16 December 2008 16:59 Go to previous messageGo to next message
Florian Uhlig is currently offline  Florian Uhlig
Messages: 424
Registered: May 2007
first-grade participant
From: *gsi.de
Hi Mikhail

You have now write permissions for the parameter directory of ecal.

Ciao

Florian
Re: ECAL parameter subdirectory [message #7643 is a reply to message #7640] Tue, 16 December 2008 17:04 Go to previous message
Mikhail Prokudin is currently offline  Mikhail Prokudin
Messages: 5
Registered: March 2006
Location: Moscow, ITEP
occasional visitor
From: 62.63.90*
Large files (12x12.root 3x3.root 6x6.root showerlib.root) can be found in /u/prokudin/temp/for_volker

I will commit the rest of files in parameters/ecal ASAP.
Previous Topic: Configuration files outside gconfig directory
Next Topic: Release DEC08
Goto Forum:
  


Current Time: Thu Dec 05 07:35:13 CET 2024

Total time taken to generate the page: 0.00888 seconds