Confluence just stops working but service is still running.

[Digikeijs]

Hi all,

It happens a few times per week and I have no idea how to solve this. Sometimes confluence just stops working but the service is still running. The only way to solve it is to stop the service and start it again. 

I thought it was my maxThreads and changed it to 60 but that doesn't help. 
Base URL, proxy, apache etc is all OK. 

I have the support file ready but not sure if I should just share it here :-)

maybe some of this would explain the issue more?

JVM Stats
--------------------------
usedMemory = 527989592
usedMemoryInMegabytes = 503
availableHeap = 8881320
freeMemoryInMegabytes = 8
allocatedHeap = 536870912
freeAllocatedHeap = 8881320
totalMemory = 536870912
totalMemoryInMegabytes = 512
availablePermGen = 0
maxPermGen = -1
maxHeap = 536870912
usedHeap = 527989592
freeMemory = 8881320
usedPermGen = -1

2018-11-04 21:05:56,358 WARN [http-nio-8091-exec-5] [common.security.jersey.XsrfResourceFilter] passesAdditionalBrowserChecks Additional XSRF checks failed for request: http://support.digikeijs.com/rest/analytics/1.0/publish/bulk , origin: null , referrer: null , credentials in request: true , allowed via CORS: false
-- url: /rest/analytics/1.0/publish/bulk | traceId: 36afcdd5ab359f10 | userName: anonymous
2018-11-04 21:06:23,362 WARN [http-nio-8091-exec-20] [common.security.jersey.XsrfResourceFilter] passesAdditionalBrowserChecks Additional XSRF checks failed for reques

... 347 more
2018-11-04 20:39:48,724 WARN [http-nio-8091-exec-11] [confluence.util.profiling.DefaultActivityMonitor] close Exceeded the threshold of 60000 ms: ActivitySnapshot{startTime=1541363901856, threadId=191, threadName='http-nio-8091-exec-11', userId='<unknown>', type='web-request', summary='/'}
-- referer: http:etc | url: / | traceId: 195f54a1e7419bbb | userName: anonymous
2018-11-04 20:40:09,053 WARN [http-nio-8091-exec-1] [confluence.util.profiling.DefaultActivityMonitor] close Exceeded the threshold of 60000 ms: A

[JP _AC Bielefeld Leader_]

Hi,

looks like the Java Heap is used up to the limit. Maybe a garbage collection of Java kicks in and makes Confluence unresponsive.

You can try switching on the garbage collection log:

https://confluence.atlassian.com/confkb/how-to-enable-garbage-collection-gc-logging-300813751.html

Maybe this gives you a hint.

Best

JP

[Tobias Anstett _K15t_]

Hi Sander,

A first point to start is to execute the log scanner in the support tools section of Confluence. This will analyse the log files for exceptions and known issues. Sometimes fixing these will already solve the problem.

It it pretty unusual that Confluence just "stops" working without anything in the logs...

If you do not kill the service, will it start responding after a certain time? If this is the case you might want to look at the threads - both http and DB. A good point to start is this article: https://confluence.atlassian.com/confkb/confluence-slows-and-times-out-during-periods-of-high-load-due-to-database-connection-pool-146407630.html

If it does not recover after a certain time you should find exceptions such as OutOfMemory or similar in the logs - the log scanner will find them.

If this does not help following up with the Atlassian support might be a good idea as sharing the support zip with them is under NDA. 

Best, Tobias