[Frost] Bank lists and disk volume for g9b cooking

Tue Oct 2 16:07:57 EDT 2012

Eugene,
trip files are not missing: since they are rather short, I added them to the 1st set of files (log+trip files).
Can we get enough space on volatile to ensure that at least the skims are kept there.

Concerning the POLfile (BOS files with limited banks): since at the end this is used in ROOT, we could run a dstmaker over it and trash the BOS file.

Adding TBER to the GPID skim is no problem but the NT_piN root ntuple has well defined entries (and doing kine.fit on single track events, is very tricky!!!)

Let's talk about it on Thursday
Franz

----- Original Message -----
From: "Eugene Pasyuk" <pasyuk at jlab.org>
To: frost at jlab.org
Sent: Tuesday, October 2, 2012 12:13:31 PM
Subject: Re: [Frost] Bank lists and disk volume for g9b cooking

First about disk space.

We have 2TB on the /work/clas/clasg9 disk. At the moment 1.5 TB is free there. We should clean it up from all pass0 stuff before going to pass1. trip file are missing on Franz's list.

Anahist files, log files, trip files, helicity files should go to the work disk. We have enough space for those.
Later on we will tar them and send to tape.

On the /volatile/clas/clasg9 we have 20TB quota and 7TB reserved. At the moment we use 8.6TB
I will request to increase reserved space up to 20TB and quota up to 25TB.

With that we should be able to write all skims and DSTs there. The largest one, full ROOTbeer DST I suggest we sent to tape and also copy on the volatile if someone wants to do something with it right away. Bu we will have to watch and delete some of them if we are running out of space. Smaller DSTs we will tar and then send to tape. While they are still on the volatile people can copy them where ever they want. 

Now about bank lists. 

We don't need to write EC1 bank because LAC was off.
For GPID skim may be it makes sense to add TGBI and TBER? 
Also may be add TBER to NT_piN. That's in case if someone would like to do kinematic fit.
Other than that it seems that everything is there.

-Eugene

----- Original Message -----
> From: "Franz Klein" <fklein at jlab.org>
> To: frost at jlab.org
> Sent: Friday, September 28, 2012 9:21:06 PM
> Subject: [Frost] Bank lists and disk volume for g9b cooking
> 
> 
> Hi,
> for the upcoming pass1 cooking of all files of g9b I compiled the
> lists
> of banks copied into output files and made an estimate of the
> required
> disk volume.
> Please make your comments and suggestions!
> Franz
> 
> (A) Disk space requirement for ca. 20,000 files (g9b runs):
> AnaHistFiles 11.6MB/file -> 230GB
> LOG-files, Trip-files ~90kB/file -> 1.9GB
> Helicity-checks  ~2.6MB/file -> 52GB (zipped -> 14GB)
> SKIMS DST_Kp 50-60MB/file -> 1.2TB
> DSTgpid 260-300MB/file ->6TB
> NT_piN 18-20MB/file->400GB
> ROOT DST 600-700MB/file ->14TB (will be copied to tape silo, not sure
> whether we can keep them on disk)
> BOS-files 4-5GB/file ->90TB (will only be kept on tape silo, not on
> disk)
> 
> 
> (B) output from cooking:
> (a) BOS output: cooked_xxxx.Axx.Bxx
> from raw data: HEAD, EPIC, TESC, HLS, TRGS, RC26, TGBI, RFT, EC1, EC,
> ECT, SC, SCT, STN0, STN1, TGTL, TGTR, TAGE (TGTL, TGTR, TAGE could be
> skipped for linearly polar. data),
> detector result banks: CL01, EC01, ECHB, SCRC, STRE, TAGR, TRKS,
> TBER,
> TDPL, DHCL, MVRT, VERT,
> particle id banks: PART TBID, GPID, HEVT, EVNT, DCPB, TRPB, ECPB,
> LCPB,
> SCPB, STPB, TGPB
> 
> (b) BOS output: Pol_xxxx.Axx (only for linearly polar. data):
> HEAD, EPIC, TESC, TRGS, RC26, RFT, TAGE, TAGR, STN0, STN1, STRE, (?
> TGTL, TGTR ?)
> 
> (c) ROOTbeer output: DST_xxxxx.Axx.Bxx.root
> from raw data: HEAD, EPIC, RC26, TGBI, RFT, EC1, EC, ECT, SC, SCT,
> STN0,
> STN1,
> result banks: CL01, EC01, ECHB, SCRC, STRE, TAGR, TBER, TDPL, MVRT,
> VERT, PART, TBID, GPID,
>       HEVT, EVNT, DCPB, TRPB, ECPB, LCPB, SCPB, STPB, TGPB
> 
> (d) SKIMS Kplus: DST_Kp_xxxxx.Axx.Bxx.root
> same as ROOTbeer DST but cuts:
> at least 1 particle with Beta>bmin_K && Beta<bmax_K && Pmom>0.1 &&
> Pmom<2.5
>   using EVNT (beta and Pmom) with
>   bmax_K= Pmom/sqrt(Pmom*Pmom+0.42*0.42)+0.05;
>   bmin_K = Pmom/sqrt(Pmom*Pmom+0.6*0.6)-0.05;
> 
> (e) SKIMS: DSTgpid_xxxxx.Axx.Bxx.root
> banks: HEAD, EPIC, GPID, TBID, ECHB, SCRC, STRE, TAGR, MVRT, VERT
> 
> (f) SKIMS: ROOTntuple NT_piN_xxxx_Axx.root
> using banks: HEAD, TGBI, EVNT, MVRT, DCPB, TAGR, SCPB, STPB
>   keeping a single pos. particle and up to 2 neutrals in a ROOTntuple
> 
> 
> 
> 
> --
> ===============================================================
>                    Franz J. Klein, Associate Professor
>                    CUA, Department of Physics
>                    Washington, DC 20064
>    office: Hannan Hall 206          phone: 202-319-6190
>    or: Jefferson Lab,CC F-243       phone: 757-269-6672
> ---------------------------------------------------------------
> 
> 
> _______________________________________________
> Frost mailing list
> Frost at jlab.org
> https://mailman.jlab.org/mailman/listinfo/frost
> 
_______________________________________________
Frost mailing list
Frost at jlab.org
https://mailman.jlab.org/mailman/listinfo/frost

-- 
===============================================================
                  Franz J. Klein, Associate Professor 
                  CUA, Department of Physics
                  Washington, DC 20064
  office: Hannan Hall 206          phone: 202-319-6190 
  or: Jefferson Lab,CC F-243       phone: 757-269-6672 
---------------------------------------------------------------