Hi,
I got an incident "Memory Pool 'Code Cache' low memory detected: 95% of 48MB used."
Where do we specify Code Cache memory or is it auto calculated based on heap memory settings?
Server RAM : 16 GB
We have set,
-XMS & -XMX 9984M
-XX:MaxPermSize & -XX:PermSize to 256M
Please suggest.
Regards,
Mahesh Deshmukh
Answer by Andreas G. ·
Hi Mahesh
I've forwarded this to the engineering team. As a first respone I got this interesting link regarding code cache: http://blogs.atlassian.com/2012/05/codecache-is-full-compiler-has-been-disabled/
I hope to get a little more insight on what that means for your dynaTrace installation
Answer by David D. ·
Hi Andreas was there any more information on this. We run into this several times and our customers potentially could be losing connections. I recieved one today:
mgtlp0011_34445_ELR@mgtlp0011: Low memory has been detected on Memory Pool 'CMS Perm Gen': 98 % of 256 MB used. |
: |
Low PermGen memory will lead to Collector crashes even if enough heap memory is available. Please check for recommended VM memory settings: Sun: -XX:MaxPermSize -XX:PermSize |
We have max permsize and permsize at 256. What are the implications of bumping this up another 256 MB? and if no implications then why would we not do that on the initial install?
Thanks,
David
Answer by Andreas G. ·
Hi Dave
I dont have more info on this - the post is already 2.5 years old and I am sorry that i didnt follow up on it.
I dont think there is any harm bumping it up by another 256MB.
That's ok Andreas! I created a ticket with the question about "any impact from increasing permgen to 512?". And the response was that this would not be a problem. We do have large cache/class cache which when we get to 6.1 will have a better ability to clean itself up. So until then, hopefully the recommendation in my ticket to go to 512MB will be sufficient.
Thanks,
David
JANUARY 15, 3:00 PM GMT / 10:00 AM ET