Cron jobs used for the Run7 Minimum Bias Data Reconstruction
Contact Person:
Charles F. Maguire
Creation Date: May 30, 2007
The database updating sequence starts with the gridFTP transfer of three
restore files from the rftpexp01 server machine at RCF. The rftpexp01
node is the only one of the rftexp nodes which is allowed by the
firebird node. The cron jobs are as follows:
- 05 0 * * * /phenix/data11/maguire/Run7/gridUpdateDataBase.csh >&
/phenix/data11/maguire/Run7/sshCommand.log &
This cron job runs at 5 minutes past midnight EDT (11:05 PM CDT) in the
maguire account at on rftexp1. It sends the three
restore files to the /mnt/eon0/rhic/dataBaseUpdating/dbRun7 of the
rhic account on the firebird node.
- 35 23 * * * /mnt/eon0/rhic/dataBaseUpdating/fdtUpdateServer.csh >& /mnt/eon0/rhic/dataBaseUpdating/fdtUpdateServer.log &
This cron job using the rhic account on the firebird node sets up an
FDT server process at 11:35 CDT. This server process will allow for FDT
transfers to the vpac04 node only.
- 35 23 * * * /rhic2/pgsql/updateRun7FromRCF/fdtUpdateClient.csh >& /rhic2/pgsql/updateRun7FromRCF/fdtUpdateClient.log &
This cron job using the postgres account on the vpac04 node sets up an
FDT client process 30 seconds after 11:35 PM CDT to copy the three restore
files from the firebird node to the local /rhic2/pgsql/dbRun7 disk area
on vpac04.
- 05 1 * * *
/phenix/data11/maguire/Run7/failureToSendDB.csh >&
/phenix/data11/maguire/Run7/failureToSendDB.log &
If the gridFTP job fails to send the three restore files to the firebird
node, then this cron job at 1:05 AM EDT using the maguire
account on rftpexp01 will mail a failure e-mail
to the maguire e-mail address at
Vanderbilt.
- #0 0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17
/var/lib/pgsql/update/startRestoreFromDumpsABC.csh >&
/var/lib/pgsql/startRestoreFromDumpsABC.log &
The above cron job is not yet operational, nor has it been tested to work.
It is a prototype of how to start the 3-cycle database update on vpac15.
Since the 3-cycle database system may be changed to vpac04 instead,
this cron job may be moved to vpac04 in the near future.
- 15 * * * * /var/lib/pgsql/monitoring/inputFireBirdOccupancyEon0.pl >& /var/lib/pgsql/monitoring/inputFireBirdOccupancyPerlEon0.log &
Script to check the disk occupancy of /mnt/eon0 on the firebird node
- 20 * * * * /var/lib/pgsql/monitoring/inputFireBirdOccupancyEon1.pl >& /var/lib/pgsql/monitoring/inputFireBirdOccupancyPerlEon1.log &
Script to check the disk occupancy of /mnt/eon1 on the firebird node
- 25 * * * * /var/lib/pgsql/monitoring/inputFireBirdNewFilesEon0.csh >& /var/lib/pgsql/monitoring/inputFireBirdNewFilesEon0.log &
Script to check for the arrival of new input PRDF files on firebird /mnt/eon0 area
- 30 * * * * /var/lib/pgsql/monitoring/inputFireBirdNewFilesEon1.csh >& /var/lib/pgsql/monitoring/inputFireBirdNewFilesEon1.log &
Script to check for the arrival of new input PRDF files on firebird /mnt/eon1 area
- 5,35 * * * * /gpfs3/RUN7PRDF/prod/run7/cautoRun7 >& /gpfs3/RUN7PRDF/prod/run7/cautoRun7.log
Script runs on the vmps18 node using the phnxreco account. The script checks
to see if the previous set of jobs has finished before launching a new set
of jobs. Various conditions are checked to see if the script should
exit immediately.
There are presently no cron jobs which monitor the progress of the
reconstruction jobs sumbmitted to PBS. Such a cron job will be
needed for the global monitoring WWW page.
- 10,40 * * * * /var/lib/pgsql/monitoring/gridFtpNanoMonitor.csh >& /var/lib/pgsql/monitoring/gridFtpNanoMonitor.log &
Script to check if the gridFTP transfer rate to RCF has dropped below 15
MBytes/second. This script runs on the vpac15 node and uses the postgres
account
- 01,31 * * * * perl -w /phenix/data11/maguire/Run7/confirmAndEraseData63.pl >& /phenix/data11/maguire/Run7/confirmAndEraseData63Perl.log &
Script to check if all the nanoDST files arrive correctly on the data63 disk at
RCF. The script runs on the rftexp01 node at RCF using the maguire
account since it uses gridFTP to send a file to the firebird node.
This script function has to be updated to be more general in terms of
knowing what is the data area at RCF which should be checked.