PS1 IPP Czar Logs for the week YYYY.MM.DD - YYYY.MM.DD

(Up to PS1 IPP Czar Logs)

Monday : 2011.10.31

• 00:30 another LAP chip

  neb://ipp025.0/gpc1/20100522/o5338g0475o/o5338g0475o.ota76.fits
        0                     NON-EXISTANT file:///data/ipp053.0/nebulous/14/50/286984446.gpc1:20100522:o5338g0475o:o5338g0475o.ota76.fits
        1 c5d2ec4d851524456fb48a1bd770d6bf file:///data/ippb01.1/nebulous/14/50/1154222719.gpc1:20100522:o5338g0475o:o5338g0475o.ota76.fits
  
  cp /data/ippb01.1/nebulous/14/50/1154222719.gpc1:20100522:o5338g0475o:o5338g0475o.ota76.fits /data/ipp053.0/nebulous/14/50/286984446.gpc1:20100522:o5338g0475o:o5338g0475o.ota76.fits
  
  --> neb-repair should have been able to fix this case but didn't in the chip stage if problem. if doesn't then need to know why.
  

• Ingestion of nebulous on ippdb02 completed during the week end (after 10 days: Wed Oct 19 16:12:33 HST 2011 - Sat Oct 29 00:09:27 HST 2011). Replication started

  CHANGE MASTER TO MASTER_HOST='ippdb00', MASTER_USER='repl_neb', MASTER_PASSWORD='xxx', MASTER_LOG_FILE='mysqld-bin.003352', MASTER_LOG_POS=277806591;
  

The slave is 1011144 seconds behind the master.

• 11:15 Mark: returning the compute2 group from deepstack back to distribution for now.
• 14:17 Bill: dropped two chips whose files have been lost

regtool -updateprocessedimfile -set_ignored -exp_id 232263 -class_id XY15
regtool -updateprocessedimfile -set_ignored -exp_id 231044 -class_id XY16

Tuesday : 2011-11-01

Heather is czar today

• 10:31 Bill had Gavin change the automount parameters to use hard instead of soft mounts. Pantasks' restarted
• 12:45 or so Found two more lost files

regtool -updateprocessedimfile -set_ignored -exp_id 187628 -class_id XY50
regtool -updateprocessedimfile -set_ignored -exp_id 187601 -class_id XY50

• chip 335562 XY67 has been faulting repeatedly. To debug Bill attempted to run it by hand and it succeeded. Strange

Wednesday : 2011-11-02

• Bill queued 20 i band M31 exposures for chip through warp processing for a demonstration of how magic makes M31.2011 data unusable.

Thursday : 2011.11.03

• 10:00 Mark adding the condor_MD07.V3_01 label to pantasks to finish the chip-warp processing for refstacks.
• 10:10 stdscience seems lagging, restarting. LAP gets held up on ipp053 jobs. Moving condor_MD07.V3_01 priority above LAP. MD07 also being held up on issues preventing jobs from completing, but is still processing other MD07 jobs.

Friday : 2011.11.04

Mark czar

• 01:05 ippc13 down, cycling power. no info on console. restarted update pantasks.
• 01:30 new data coming in. stalled. restarted registration and summitcopy. was hung mount on ipp053 to ippb01, cleared with

  /usr/local/sbin/force.umount ippb01
  

• 07:50 ippc09 mount of ippb02 stuck.
• 08:00 shutdown of stdscience pantasks dumped backtrace to ippc16 stdout

  ipp@ippc16:/home/panstarrs/ipp>*** glibc detected *** pcontrol: double free or corruption (!prev): 0x00000000006d7730 ***
  ======= Backtrace: =========
  /lib/libc.so.6[0x7fc41c95bb88]
  /lib/libc.so.6(cfree+0x76)[0x7fc41c95d746]
  pcontrol[0x406571]
  pcontrol[0x406b68]
  pcontrol[0x406bac]
  pcontrol[0x403b6e]
  /home/panstarrs/ipp/psconfig//ipp-20110622.lin64/lib/libbasiccmd.so(quit+0x18)[0x7fc41e69fba8]
  /home/panstarrs/ipp/psconfig//ipp-20110622.lin64/lib/libshell.so(command+0x22c)[0x7fc41e47cc5c]
  /home/panstarrs/ipp/psconfig//ipp-20110622.lin64/lib/libshell.so(multicommand+0x90)[0x7fc41e480cd0]
  /home/panstarrs/ipp/psconfig//ipp-20110622.lin64/lib/libshell.so(opihi+0x90)[0x7fc41e493b30]
  pcontrol[0x403bd9]
  /lib/libc.so.6(__libc_start_main+0xe6)[0x7fc41c905486]
  pcontrol[0x401f69]
  ======= Memory map: ========
  00400000-0040d000 r-xp 00000000 00:14 17252593                           /data/ippc18.0/home/ipp/psconfig/ipp-20110622.lin64/bin/pcontrol
  ...(full map in notes)...
  
  

• 09:10 Bill set states for 2 lost rawImfiles

chiptool -dropprocessedimfile -set_quality 42 -chip_id 337393 -class_id XY53
regtool -updateprocessedimfile -set_ignored -exp_id 195736 -class_id XY53

chiptool -dropprocessedimfile -chip_id 337396 -class_id XY37 -set_quality 42
regtool -updateprocessedimfile -exp_id 195761 -class_id XY37 -set_ignored

• 11:30 Mark stopping stdscience, distribution while manually fixing all hanging mounts. ipp053 seems to have particular trouble with ATRC mounts, removing from processing in adding to the hosts_ignore_wave3 group.
• 12:50 all stalled/hanging processes and ippbXX mounts cleared. restarting stdscience and distribution.
• 13:00 many jobs want to use ipp053 (ota76), so will take some time to clear
• 13:10 chip_imfile taking >1000s on ipp014, neb-repair stalled, needed to reset mounts to ippb01.
• 14:15 Bill asked Gavin to re-set the ippbXX mounts to soft mounts as before (-nosuid,rw,tcp,soft,rsize=32768,wsize=32768,timeo=20,retrans=6,retry=5)
• 14:42 Bill set to diffRuns with no diffInputSkyfiles to drop 186382 & 186805
• 14:46 Bill set 11 LAP destreak runs in state 'failed_revert' to 'new'
• 15:00 last nights 65 exposures finally though. MD07 has many warps that also want to run on ipp053, LAP set mostly cleared of remaining ipp053 need.
• 14:30 some LAP exposures still waiting for update since 11/2, unsuccessful in getting to update so dropping exposures so stack can finish

  laptool -updateexp -lap_id 1593 -exp_id 384384 -set_data_state drop -dbname gpc1
  laptool -updateexp -lap_id 1593 -exp_id 384366 -set_data_state drop -dbname gpc1
  laptool -updateexp -lap_id 1593 -exp_id 235422 -set_data_state drop -dbname gpc1
  laptool -updateexp -lap_id 1593 -exp_id 102133 -set_data_state drop -dbname gpc1
  

• 20:30 ippc13 down, rebooting. no info on console, load was above normal, may have overloaded

Saturday : 2011.11.05

Sunday : 2011.11.06