Hello,
Performance troubleshooting is something that is going to require a lot of research to pin down the actual cause as there are so many things that could be a factor, from the communication at the network, to the application configurations or even the hardware causing a hiccup.
I would recomend starting with this KB we have on the topic "Troubleshooting Performance Problems", noting the first paragraph of the Overview:
Troubleshooting Performance Problems is not a trivial process and can be quite complicated when it comes to identifying problems with JIRA applications. Given the number of different configurations possible for JIRA applications, such as different versions, plugins, network, hardware configurations and other various factors, it increases the difficulty of solving these problems. This Knowledge Base article is designed to assist in making that job easier to do and also provide guidance in identifying problems and also giving information to Support.
And also look into the following articles as well, which are collections of various documentation material to help narrow down the scope of performance issues:
One point to look at would be to check for garbage collection errors in the logging, because looking at the memory profile you listed noting ~7.9GB allocated but only 450 issues total which seems like a lot of memory for a smaller instance, but this is entirely contingent on the other factors within the instance such as add-ons, issue type count, custom field count, project count, etc... looking at the Jira Server Sizing Guide as a referance point for memory usage expectations based on instance size, you are between Small and Medium scale based on issue count alone excluding all other factors the ">1GB" but less than 2GB would be the target memory range on HEAP until your instance grew to the larger scale size.
Noting Too much memory can be an issue itself. As There could be too much memory allocated to the instance HEAP space triggering additional garbage collection overhead, Some more details on this can be seen in the article "JIRA applications crash due to OutOfMemoryError GC Overhead Limit Exceeded" noting the application is not fully crashing but the comment at point 4.a under the resolution sums this one up:
When allocating memory to a JVM, more memory does not always equate to a better experience. If there is a problem with third-party plugins or JIRA applications are running into certain behaviors that is causing it to use memory at a drastically increased rate, applying additional memory to the JVM can have detrimental effects and often addresses the symptom rather than the root cause. An increase in heap size will lead to longer GC times, which can render instances "frozen" when GC occurs, which can sometimes take up to 10 seconds (worse in certain cases).
And while memory settings could be a good starting point this may or may not be related at all because a complete contradictory statement to the above can be seen in the "Garbage Collection (GC) Tuning Guide" noting in certain circumstances the "More memory the better".
Overall its a delicate balancing act between all the various parts, and you will need to isolate the parts one at a time until more data can be collected on the choke point, check the logs and thread dumps first let me know what you find and depending on the timetable your working with an additional recommendation to speed things up for performance tuning help would be to look into hiring an Atlassian Expert partner on contract that specializes in application performance tuning that could be contracted in to make recommendations for your specific setup on a increased timetable.
Regards,
Earl
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.