Tomcat performance issues

vish vish July 31, 2012

Hi,

Jira is being hosted by one of our servers which is a well equipped with 16GB RAM and 200 GB disk space. Tomcat seems to be using a lot of CPU from time to time. CPU utilization increases to 100% randomly and falls back below 50% then. This has caused Jira application to be unavailable and is a lot of incovenience for our users.

Currently this is the only application hosted on the server and still Jira is not stablized yet. We are planning to host another application on the server sometime soon but we are weary of this as Jira alone is not working well with the server.

Please suggest any improvements/solution to this issue. Are there any configuration changes available to improve permofrmance of tomcat and also if there are any logs produced by tomcat.

Thanks,

Vishakha

2 answers

1 accepted

0 votes
Answer accepted
EddieW
Rising Star
Rising Star
Rising Stars are recognized for providing high-quality answers to other users. Rising Stars receive a certificate of achievement and are on the path to becoming Community Leaders.
August 1, 2012

Follow Jira performance guidelines outline here https://confluence.atlassian.com/display/JIRAKB/JIRA+Performance+Tuning.

We run Bamboo, Jira, COnfluence, Crowd, and FIsheye on a single server, 4CPU, 8Gigs of ram without much impact to users.

How ram did you allocate to Jira? 16 gigs on the box means nothing if jira can only access the default (768MB ?) amount. And exeecisve Garbage collection due to low memory can strain CPU.

You can also crank up some logging to understand what the server is doing while the CPU is pegged (we had a google search appliance really beat up our server).

ALso, do not share multiple applications on the same tomcat instance. Tomcat has very low overhead to spin up multiple instances (especially considering a 16gig box) and Atlassian discourages using the same instance for multiple apps.

0 votes
vish vish August 3, 2012

Thanks Eddie for your answer. The information provided is going to be very helpful.

I am exploring the system and the configuration to determine if that is the culprit. We haven't changed anything since we upgraded.

Here is what I got from our Jira instance. I guess it will be helpful to assign some more memory. Please let me know your thoughts.

Thanks!

<th colspan="2">Java VM Memory Statitistics</th><th>Total Memory</th> <th>Used Memory</th> <th>Total PermGen Memory</th> <th>Free PermGen Memory</th> <th>Used PermGen Memory</th> <th>Free Memory</th>
474 MB
381 MB
256 MB
124 MB
131 MB
93 MB
<th colspan="2">Memory Pool Info</th>
PS Old Gen: init = 89522176(87424K) used = 321209440(313681K) committed = 357957632(349568K) max = 357957632(349568K)
PS Perm Gen: init = 16777216(16384K) used = 137493416(134270K) committed = 137756672(134528K) max = 268435456(262144K)
PS Survivor Space: init = 5570560(5440K) used = 27329320(26688K) committed = 35717120(34880K) max = 35717120(34880K)
PS Eden Space: init = 33554432(32768K) used = 50699392(49511K) committed = 103546880(101120K) max = 105512960(103040K)
Code Cache: init = 2359296(2304K) used = 32662464(31896K) committed = 32866304(32096K) max = 50331648(49152K)

Suggest an answer

Log in or Sign up to answer