Hi,
Our JIRA instance repeatedly fails to allocate memory, which forces a garbage collection ( see log extract below ). This doesn't seem right, does anyone know if this is expected behaviour or how if could be rectified?
2018-10-11T08:51:15.350+0100: 2391691.952: [GC (Allocation Failure) [PSYoungGen: 674740K->12516K(681472K)] 1711894K->1050236K(1724416K), 0.0524143 secs] [Times: user=0.07 sys=0.01, r
eal=0.05 secs]
2018-10-11T08:51:19.777+0100: 2391696.379: [GC (Allocation Failure) [PSYoungGen: 677604K->11170K(681472K)] 1715324K->1050424K(1724416K), 0.0597589 secs] [Times: user=0.07 sys=0.01, r
eal=0.06 secs]
2018-10-11T08:51:24.986+0100: 2391701.588: [GC (Allocation Failure) [PSYoungGen: 676258K->10053K(681472K)] 1715512K->1050237K(1724416K), 0.0491620 secs] [Times: user=0.06 sys=0.00, r
eal=0.05 secs]
2018-10-11T08:51:29.511+0100: 2391706.113: [GC (Allocation Failure) [PSYoungGen: 676165K->13181K(681984K)] 1716349K->1053950K(1724928K), 0.0533946 secs] [Times: user=0.07 sys=0.01, r
eal=0.05 secs]
Thanks
Simon
Hi @Simon Logan,
you're right, GC should not be that often. Can you please share some information about the server hosting Jira, specifically the amount of RAM ?
It would be useful to also have a look at what is set for the JVM memory settings in <jira_install_dir>/bin/setenv.sh at the JVM_MINIMUM_MEMORY and JVM_MAXIMUM_MEMORY setting.
You can also have a look at the Atlassian documentation on JVM settings here.
Let me know if this helps,
--Alexis
Hi Alexis,
Thanks for the response.
JVM_MINIMUM_MEMORY="384m"
JVM_MAXIMUM_MEMORY="2048m"
free -m
total used free shared buff/cache available
Mem: 3789 2866 192 32 731 593
Swap: 2047 775 1272
Does that give you what you need?
Thanks,
Simon
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi @Simon Logan,
I have a similarly sized server and the following settings are working fine for me:
JVM_MINIMUM_MEMORY="384m"
JVM_MAXIMUM_MEMORY="1024m"
How big is your Jira instance in terms of number of users, projects, issues ... ?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
@Simon Logan I can see this is a fairly big instance, I'd suggest having a look at this page that will provide some pointers on how to size your server memory and JVM.
Based on Atlassian recommandations, I'd suggest bumping your memory from 4GB to 8GB at least and setting the heap size to 4GB.
Let me know if this helps,
--Alexis
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi,
if you go for 4GB (espessaly 6+GB) you should switch also from paralelgc to g1gc. (set-gc-params.sh/bat).
BR
Udo
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.