powered by Jive Software

Openfire Out of Heap memory

Hi Team,

I am using open-fire server 4.1.3 on Ubuntu machine with default heap memory 239.75 MB.

We do have registered only 15 messages and and total roster entry is 141.

We have created 44 Roster Group for exchanging message in group.

Now if we start the server after a 2 day we got java heap out of memory so we tried to increase it to 512 mb but after increasing we face another problem like service is sacrificed as server doesn’t have enough memory so it is killing our service without notifying us.

We hardly exchange few messages not more then 1k message per day still we are facing this issue so what is best way to handle this case.

Does anyone come across this issue then please help us to move further.

Thanks,
Hiren Vavadiya

Not sure what sacrificing means. If this is some mechanism of OS (Ubuntu), then you should look into OS options to override or disable such behavior. Or maybe add more RAM to that server.

Thanks for you replay,

I will look into os configurable parameters.

how much ram you are suggesting for 15 users,150 roster and 50 roster groups?

15 users doesn’t seem much, but i don’t know your usage pattern and why you have so many groups and rosters with just 15 users. Add 1 GB and check how it goes. If still problematic, add more and so on. Increasing Openfire’s max heap while adding RAM also could be useful.

We are testing our application so we have created many group which never deleted.

each users have 10 user in his roster list so the figure is very near to 150.

we are exchanging around 1k message each day and some time it could be very near to 5k to test our application.

you are suggesting increasing RAM and heap so do you think memory leak still exists in openfire server version 4.1.3 ??

Thanks

I don’t know if it is a memory leak or just a normal memory load under such usage. Of course, updating to the recent version is recommended. As i’ve said, i would add more RAM in increments, as well as adding more to max java memory value and see when it stops failing. Btw, you would have to switch to 64-bit version to add more than 2 GB to java memory.

thanks for your replay.

we are working on 64 bit binary and we know increasing heap memory we will try to do the same you suggested and will get back to you after my observation.