[ome-users] Logging levels

[Josh Moore]

Hi John,

I've listed some of the most numerous log statements from one of our
production systems on the ticket:

  https://trac.openmicroscopy.org.uk/omero/ticket/2138

Of those, the most important for knowing what's going on on your
system are ServiceHandler and EventHandler. They are the equivalent of
access.log for Apache.

If you want to generally quiet your logging, I'd suggest setting the
threshold setting at the top to WARN:

    <param name="Threshold" value="DEBUG"/>

becomes

    <param name="Threshold" value="WARN"/>

This doesn't require a restart since log4j will notice the file
change. Before sending bug reports, though, it will usually be helpful
if not necessary to turn this back to DEBUG or INFO for us to know
what's going on.

If anyone has an optimized log4j.xml or knows of performance issues
from server logging, we'd also be very interested to hear.

Cheers,
~Josh

John Wallach writes:
 > Greetings,
 > 
 > In 4.1.1, log4j is set to INFO for several ome classes that make for
 > rather large log files. Specifically:
 >   <category name="ome"> <priority value="INFO"/> </category>
 >   <category name="ome.formats"> <priority value="INFO"/> </category>
 >   <category name="omero"> <priority value="INFO"/> </category>
 >   <category name="OMERO"> <priority value="INFO"/> </category>
 > 
 > In the size of TB hard drives the size of the log files isn't that
 > worrisome but the time that the app spends writing out all of this
 > concerns me. Has anybody turned these down to WARN or even ERROR?
 > And if so does it work :)?
 > 
 > Thanks,
 > jmw