Outof memeory errors

Hey guys, I understand that basis, however I wanted to know, why am I getting out of memeory errors when I have this much RAM left? 229.73 MB of 1637.06 MB I have 1400MB of ram remeaining, yet still it shows OOM errors in the logs,. Everytime I have to reboot the server due the connection to gmail and my legacy server ‘‘locking up’’ I go through this issue. I ahve Xms 512 and Xmx 1700, yet still it isn;t enough. any more clues at what I can do to optimize this thing?


reset the stack size to 256k and it began accepting connections again. Not the way I wanted things to turn out. For some reason when 800+ connections are hitting the server at the same time, it renders the server useless. will not allow connections, and just hangs. Luckily found the post on the stack size. Think next time I reboot, I will try the 128k suggestion on that same post. For those of you that are wondering -“Xms512M -Xmx1500 -XX:+UseParallelGC -Xss256k” Do not know if that is the best setup, but It works for me.

Message was edited by: jeff_garner

Hey Jeff,

I’'m confused. Is your server frozen and not accepting new connections? Or is it just an OOM error that is being printed but the JVM is still working (somehow).

FYI, when using the old ssl method (port 5223) in Wildfire 3.2.0 then you may potentially end up with a blocked JVM. That problem has been fixed for Wildfire 3.2.1. The nightly build version already includes the bug fix.


– Gato

Hi Jeff,

which kind of OOM errors do you see? There are native and heap ones.


I had already closed this prior to getting your answers. started a new thread. Seems that I find a fix, then I get another 200 or 300 more users out of it. I am up to 1050+ users, but then I start getting the OOM errors again. to answer Gato’'s question, I see the unable to create connection error, but the other people on the server are communicating, although the PyMSN client seems to stop communicating, and a TON of LDAP connections are open at all times… very very low allocated mem usage… I am confused. Check out my other post for newer info.