|
This is working for my case (using the standalone build), perhaps this solution is only valid for the standalone versions of jira, but it really would be nice. I would note that this is how confluence is setup by default, so it may be worth changing in jira as well.
John,
Thanks for the update. We will take a look. Cheers, John,
You are right, we can get it working for JIRA Standalone. Cheers, Anton | John:
To improve performance, use the storage report numbers generated by the RPTSTG run-time option as an aid in setting the initial and increment size for STACK. One: Hello Anton
One cool feature of log4j is the posibilty of reading variables from command line, you can do stuff like: Define a log4j.properties or .xml if its the case and do: log4j.properties: (tomcat example) And it will pick the loglevel from the defined variable. Changing the logfile path in log4j.properties may work, but it has an unwanted side effect.
When I start a Support Request the old location of the logfile is used, even after a jira restart. Also when you look at the entry "Location of atlassian-jira.log" in "System Info" you get the old location. This means, log4j puts its logfiles in the new path, but jira does not know about it! Cheers |
|||||||||||||||||||||||||||||||||||||||||||||||||||
Does ${catalina.home}/logs/atlassian-jira.log actually work? I think log4.j does not interpolate the ${catalina.home} variable as it does not know if it is running in Tomcat.
Please see Jeff's comment for the explanation:
http://jira.atlassian.com/browse/JRA-9769#action_77920
We would like to fix this problem once we implement JRA-1642.
I will resolve this issue for now. Please let me know if I missed the point and we will reopen it.
Cheers,
Anton