wiki:logJMD

Version 1 (modified by niles, 10 years ago) (diff)

--

JMD Log Files

As already discussed, a JMD log file will have a name similar to "2014_02_10.NSO.stderrout.log" and will log successful downloads with lines like :

Feb 10, 2014 1:12:38 AM org.vso.jmd.Downloader.SCPDownloader call INFO: Th ID:[9254]; SU:[528482910]; RN:[131877503];[aia.lev1];[MIRROR]; Sz:[11275255]; STP:[DONE]; ST:[DONE] [611.72174KB/s] [SAO]

Failures will be recorded with lines like :

Feb 10, 2014 12:07:25 AM org.vso.jmd.Downloader.SCPDownloader call INFO: Th ID:[9235]; SU:[528417849]; RN:[131843041];[aia.lev1];[MIRROR]; Sz:[7626390]; STP:[SCP]; ST:[FAIL] [N/A] [JSOC]

Thus, a simple failure and success counts can be achieved with :

grep FAIL 2014_02_10.NSO.stderrout.log | wc -l
grep DONE 2014_02_10.NSO.stderrout.log | wc -l

However, it may be more useful to break the failure/success counts down by site, with a bash script like so :

jmdLogDir=/my/jmd/log/dir
jmdLogFile=`/bin/ls -C1 $jmdLogDir/*.stderrout.log | sort | tail -2 | head -1`
echo Yesterday\'s JMD downloads according to $jmdLogFile
grep '\[DONE\]' $jmdLogFile | cut -d\; -f8 | cut -d" " -f4 | sort | uniq -c
echo Yesterday\'s failed JMD downloads :
grep '\[FAIL\]' $jmdLogFile | cut -d\; -f8 | cut -d" " -f4 | sort | uniq -c
echo

which will generate output similar to this :

Yesterday's JMD downloads according to /opt/JMD/log/2014_02_04.NSO.stderrout.log
  66564 [JSOC]
  31356 [SAO]
Yesterday's failed JMD downloads :
      1 [JSOC]