hydrophone

After deployment of a new hydrophone

• copy the ssh keys using ssh-copy-id, something like gcouser@gconode01:~$ ssh-copy-id '[email protected]'
• copy the cross-compiled rsync file to the hydrophone

gcouser@gconode01:~$  ssh '[email protected]' mkdir -p bin
gcouser@gconode01:~$  scp -p path/to/cross-compiled/rsync [email protected]:bin/
gcouser@gconode01:~$  ssh '[email protected]' bin/rsync --help

After a redeployment

• Update hydrophone_config.csv
• Then power cycle the device.

Audio data problems...

• If the python receiver cannot be reached...
  • spdivid server might be down?
  • is the collect_hydrophone app showing ok in supervisord?
    • sudo supervisorctl status
  • there might be a problem with supervisord on spidvid?
  • there could be another problem on spidvid
  • There may be a network problem?
• If no data in the python receiver
  • is it connecting to the correct moxa switch and port (configured in the hydrophone_config.csv file)
  • Is the hydrophone configured (using Lucy or the web application) to stream audio data?
• If collector running ok, but no data in the archives
  • check the crontab and logs on dmdmznexstor01 vm

opsuser@dmdmznexstor01:~$ crontab -l | grep audio
 * * * * * /home/opsuser/bin/rsync-audio.sh >> /home/opsuser/logs/rsync-audio.log 2>&1 || exit 0

FFT data problems...

• Check the log file on gconode01 /home/gcouser/logs/hydrophone.log
  • Is the hydrophone configured (using Lucy or the web application) to stream collect fft files?
  • Does the crontab on gconode01 look OK?

gcouser@gconode01:~$ crontab -l | grep hydrophone
* * * * * /home/gcouser/dev/uwobs/common/apps/collectors/collect_hydrophone.sh >> /home/gcouser/logs/hydrophone.log 2>&1

• perhaps it is not a hydrophone specific problem?

  • could be a problem with the kafka cluster?
  • could be a problem with mirrormaker on kafka02
  • could be a problem with the small kafka (in docker) node on gconode01?
  • could be a full disk or some other problem on gconode01

• Are the FFT data files appearing in the archive ?

  • Is the cron job running OK on dockerub (spiddal.marine.ie) vm?

dmuser@dockerub:~$ crontab -l | grep hydro
*/1 * * * * rsync -avz --remove-source-files --include '*.txt' --exclude '*' -e ssh [email protected]:/home/gcouser/hydrophone/Data/ /home/dmuser/incoming hydrophones
*/1 * * * * sleep 10; for item in $(ls -1 /home/dmuser/incoming/hydrophones/*.txt); do target=/home/dmuser/sites/spiddal.marine.ie/html/data/hydrophones/$(echo $item| sed -e 's#^.*/\(\(.......\)_\(....\)\(..\)\(..\)\(.*\)$\)#\2/\3/\4/\5/#') && mkdir -p $target && mv $item $target ; done

• if data archive seems ok, but no data in erddap
  • if the data is showing in mqtt.marine.ie (give a minute for the data to show) then it's not a problem with kafka.
  • could be a problem with cassandra
  • could be a problem in node-red