[Frost] Bank lists and disk volume for g9b cooking
Eugene Pasyuk
pasyuk at jlab.org
Tue Oct 2 16:43:37 EDT 2012
Franz,
I requested more space on volatile, 25TB quota and 20TB to be reserved. If we get this it should be enough for all skims.
-Eugene
----- Original Message -----
> From: "Franz Klein" <fklein at jlab.org>
> To: "Eugene Pasyuk" <pasyuk at jlab.org>
> Cc: frost at jlab.org
> Sent: Tuesday, October 2, 2012 4:07:57 PM
> Subject: Re: [Frost] Bank lists and disk volume for g9b cooking
>
> 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
> ---------------------------------------------------------------
>
>
More information about the Frost
mailing list