Project

General

Profile

Restart WEB OM processes

(updated by DT on 1/22/2018)

If you notice that the either ND or FD OM Web page "Current run:" does not match the one shown in the RcGUI, you'll need to restart the WEB OM processes as follows:

Note: ND/FD OM run on minos-nearline

1) Login: ssh -X
2) Use "ps -aux | grep loon" to find and kill the version of "loon -b -q macros/plotOnlineMonitoringNearLive.C" (or plotOnlineMonitoringFarLive.C) that is running
Make sure it's plotOnlineMonitoringNearLive.C and not plotOnlineMonitoringNear.C (or Far) - that's the keepup-style reprocessing of completed runs that doesn't skip any records and makes sure all subruns are plotted in case the OM got restarted halfway through a run.
3) for ND OM: cd /minos/app/home/mindata/OnlineMonitoringSoftware/releases/R305

3') for FD OM: cd /minos/app/home/mindata/OnlineMonitoringSoftware/releases/R305_FD

4) rm OMRunning.tmp (if it exists)
5) . manual_restart_om.sh (for ND)

IF the process keeps dying, please check the log file
/minos/app/home/mindata/OnlineMonitoringSoftware/releases/R305/OMLOGFILE.log

if it says
mkdir: cannot create directory `/var/tmp/mindata/mindata_lock_live_nd_om': File exists
Mon Jan 22 13:32:56 CST 2018: Couldn't start live ND OM script - lock directory exists
This means the script is already running, so this process will exit

then you must remove the lock file:
cd /var/tmp/mindata
rmdir mindata_lock_live_nd_om

and then try again restarting the process:
. manual_restart_om.sh

You should see the process running again, as for example
mindata@minos-nearline[~/OnlineMonitoringSoftware/releases/R305] $ ps -aux | grep loon
mindata 21380 39.3 3.2 267064 196204 pts/1 S 13:54 0:05 loon -b -q macros/plotOnlineMonitoringNearLive.C

After the process starts, wait ~5-10 minutes.
At this point, the "Current run:" should match the one shown in RcGUI.