One application process disappeared after other one started

[Marvin]

Hi,

I have an issue on applications.

I deployed JIRA 6 and confluence 5.1.4 on my VPS, after I restart the service, the java process seems like they are killing each other, after the JIRA start up, the confluence's java process will disappeared, after I started up confluence, JIRA's disappeared, and I have monitoring the two applications' catalina.out, everything is smoothly, just like there are some one killed one java.

can anyone help me?

[jing_hwa_cheok · Accepted solution]

Hi Marvin,

Could it be possible that they are not killing each other just overwriting each other cookie? Because if both JIRA and Confluence have the same context path, their cookies will overwrite each other therefore creating an illusion that they are killing each other session.

Browser cookies does not take into account the port therefore when you log into Confluence after you log into JIRA, your JIRA cookie will be overwritten and therefore you log out automatically. By default both of the application context path is "/"

Change the context path in the server.xml file for both Confluence and JIRA and them restart them.
You can refer to this document for more detailed instruction https://confluence.atlassian.com/display/DOC/Configuring+the+Server+Base+URL.

Hope this helps.

[Marvin]

Thank you for your suggestion, I'm already changed the path, let's take a look

[Marvin]

My JIRA still disappeared, the last log event is:

Jul 04, 2013 1:34:36 PM org.apache.jasper.compiler.TldLocationsCache tldScanJar

INFO: At least one JAR was scanned for TLDs yet contained no TLDs. Enable debug logging for this logger for a complete list of JARs that were scanned but no TLDs were found in them. Skipping unneeded JARs during scanning can improve startup time and JSP compilation time.

I've already add /jira and /confluence path

[jing_hwa_cheok]

Hi Marvin,

This issue seem to be quite complicated. I believe it is best for you to create an issue at support.atlassian.com and our engineers will look into this case with you.

Also generate the full support zip (Administration > System > Atlassian Support Tool > Support Zip) to get faster response.

Cheers,

Jing Hwa

[Renjith Pillai]

If you are already seeing Java OOM error, then you do not have a choice but to increase the server memory. Keep JIRA and Confluence on 512MB if you are having a small instance.

[Marvin]

I find out who killed my java:

please check my attached kernel message (message.txt), My VPS have 2 GB RAM, but it always java oom, can anyone help me?

[Marvin]

I find out who killed my java:

please check my attached kernel message , My VPS have 2 GB RAM, but it always java oom, can anyone help me?

[Jobin Kuruvilla [Adaptavist]]

Anything in the other logs? lovalhost logs or application logs?

[Marvin]

only confluence have some error:

ERROR [scheduler_Worker-1] [atlassian.crowd.directory.DbCachingDirectoryPoller] pollChanges Error occurr

ed while refreshing the cache for directory [ 131073 ].

But arround it I can't see any about java process event