Last Updated on Monday, July 30, 2001 at 21:33:43. by Author: Burt Holzman
1. Ctrl-C or kill daq_daemon, ped_daemon, deletor_daemon, PedSelect (in that
order)
2. If we have time, move data on vmesparc to olddata subdir
3.
on phucsia:
phat
f = new TPhPedSelect()
ssh phobos@phobosx.phobos.bnl.gov,
make sure RSA keys are ok
cd datamover/src
./daq_daemon -nd -nr
ssh shift@phast.phobos.bnl.gov
cd datamover/src
./ped_daemon
4. Make sure daq_daemon is caught up re-sinking if there are files currently on
disk. Shift crew can confirm this.
5. DAQ may still be locking a partition. Start a new run, or wait until
new sequence (or stop current run if we are taking junk)
6. ssh to vmesparc. Check latest files on disk and confirm they have been
sunk.
7. On phobosx:
ps -ef | grep daq_daemon
kill -16 [PID] starts up recon mode
8. Run ~phobsink/bin/write_and_recon_nodel.bnh from phobosx on any runs that
have been sunk in -nr mode
but not reconstructed. Check /data/7/datamover/readlock and run the script
right after the daq_daemon locks the partition -- otherwise it might unmount it
before the script is run