Re: [Brahms-dev-l] brahms disk clean up

From: Hironori Ito <hito@rcf.rhic.bnl.gov>
Date: Thu Sep 02 2004 - 17:58:36 EDT
Hello.  Here is the list of directories hogging NFS disks.  Or, in other 
words, this is the list of directories I would delete at first. (very 
soon if you don't say anything.)  As you can see the total wasted(???) 
space is very large (several TB).  Please check this list if you don't 
want to loose these directories.

This is my suggested course of actions to these directories.

1.  For ltr,gtr,dst files for run01---run03, if they are used for paper, 
rename them to be official files.  And, they can be moved to HPSS.  This 
requires someone who made those files to declare that they are official 
files.

2.  For ltr, gtr, dst files for run04, just delete them if they are not 
made by official script.

3.  For personal files, if they are semi-permanent and very large, they 
should be moved to HPSS (by user with hsi and htar).  Small files can be 
in NFS disks, but remember that NFS disks are not backed up.  People can 
always use local disks (/home), and there are alot of them (every rcrs 
and rcas machines).



/brahms/data01/ejkim                          ----> 193 GB ( mrs map for 
run01).
/brahms/data01/reduced/R5XXX        ----> 324 GB ( sequence files from 
run01)
/brahms/data01/reduced00                   ----> 62GB ( sequence files 
from run00)
/brahms/data01/tofcal00                       ----> 90GB ( sequence 
files from run00)
/brahms/data02/data                              ----> 530GB ( data from 
Run02, Kris declared that these files can be moved to HPSS.)
/brahms/data02/igb/2003                      ----> 168GB ( test files 
for run03)
/brahms/data02/reduced                        ----> 195GB ( sequences 
files from run02)
/brahms/data03/jij                                 ----> 407GB 
(acceptance simulations for run02)
/brahms/data03/tracktree                      ----> 109GB (tree files 
from run02)
/brahms/data03/reduced                        ---->  92GB ( sequecs 
files from run02)
/brahms/data04/data/gtr                        ----> 194GB (gtrs from 
data02)
/brahms/data04/reduced                        ----> 491GB (seq files 
from run02)
/brahms/data05/data/gtr                        ----> 263GB (seq files 
from run02)
/brahms/data05/jij                                  ----> 231GB 
(acceptance simulations for run02)
/brahms/data05/gtr                                 ----> 155GB (gtr 
files from run03)
/brahms/data06/data/gtr                         ----> 170GB ( files from 
run02)
/brahms/data06/ejkim                             ----> 229GB ( non 
official gtr from run03 and run04)
/brahms/data06/cholm                           ----> 291GB (analysis 
files from run02
/brahms/data07/scratch/bjornhs            ----> 307GB ( non official 
files from run03 )
/brahms/data07/scratch/gtr                   ----> 172GB (non official 
gtr files from run03)
/brahms/data08/data/run04                   ----> 390GB (non official 
files from run04)
/brahms/data08/scratch/bjornhs            ----> 116GB (analysis files 
for run03)
/brahms/data08/jij                                 ----> 631GB (analysis 
files for run02)
/brahms/data09/data/gtr                       ----> 46GB (files for run02)
/brahms/data09/data/pp                        ----> 34 GB (files for run02)
/brahms/data09/jij                                 ----> 199GB (analysis 
files for 199GB)
/brahms/data09/scratch/bjornhs            ----> 97GB (analysis files?? 
for run02)
/brahms/data09/data/pp                        -----> 94GB (sequence 
files for run02)
/brahms/data10/ejkim/auau                  -----> 267 GB (non official 
gtr files for run04)
/brahms/data10/pid04                           -----> 156GB (pid 
files??? for run04)
/brahms/data10/bjornhs                        -----> 146GB ( pp analysis 
files for run02)
/brahms/data11/gtr                               -----> 82GB (non 
official gtr files for run04)
/brahms/data11/jij                                 -----> 252GB 
(acceptance map)
/brahms/data12/data/auau/gtr               -----> 268 GB (files from run02)
/brahms/data12/pid03                           -----> 65GB (pid files 
for run03)
/brahms/data12/jij                                 -----> 85GB ( simulation)
/brahms/data13/ltr04                            -----> 378GB (not 
official ltr files)
/brahms/data14/jij                                 -----> 86GB (simulation)
/brahms/data15/scratch/dau                 ------> 115GB (run03 gtr files)
/brahms/data16/scratch/ejkim              -----> 427GB (not official gtr 
files for run04)


Hiro


Hironori Ito wrote:

> Hello.  This is to inform you that we need to move our data files 
> (particularly those from run01, run02 and run03) in NFS disks 
> (/brahms/dataXX)  to HPSS.  I have already put run03 data (from data07 
> and data09) to our FILECATALOG database, and I will copy those data 
> files to HPSS (within 24hours).   After that, those files will be 
> deleted from our NFS disks.  If you know locations of official data 
> files ( ltr, gtr and dst for run01 and run02) (particularly those used 
> for our paper, please inform me as soon as possible (hopefully in a 
> few days). (We may also keep calibration and a few other files.)   I 
> will rename them to our official standard format and put them in our 
> database.  After that, I will copy them to HPSS and delete them from 
> the NFS disks.  This is very crucial since we don't have enough space 
> to analyze run04 data.  Also, it is very important to inform me as 
> soon as you can since I will start to delete any not-official files 
> from NFS disks sometime soon.     If you want to keep your own files, 
> copy them to HPSS!!!  Also, there is a plenty of local disk space in 
> each machine (/home).  Although local disks are not cross-mounted in 
> the different machines, you can use ROOT daemon (rootd and TNetFile) 
> to access them from different machines.   (NOTE:  /home disks are not 
> necessary parmanent places.  It is possible that files would be 
> deleted during the RCF upgrade once a year. The next upgrade is coming 
> very soon.)
>
> Hiro
>
>
>
> _______________________________________________
> Brahms-dev-l mailing list
> Brahms-dev-l@lists.bnl.gov
> http://lists.bnl.gov/mailman/listinfo/brahms-dev-l



_______________________________________________
Brahms-dev-l mailing list
Brahms-dev-l@lists.bnl.gov
http://lists.bnl.gov/mailman/listinfo/brahms-dev-l
Received on Thu Sep 2 17:58:57 2004

This archive was generated by hypermail 2.1.8 : Thu Sep 02 2004 - 17:59:19 EDT