Win a copy of The Java Performance Companion this week in the Performance forum!
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic

Server unresponsive: memory issue or cluster issue?

 
Sujoy Choudhury
Ranch Hand
Posts: 136
Eclipse IDE Ubuntu
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Hi All,
We have 24 managed instance server in a cluster, each one having 4GB/4GB (min/max) memory settings.
We are using Weblogic 10.3 with jRockit28 (Java 1.6) on Linux.
We have setup the ClusterAddress and Listen Address settings properly.

Still when we run a certrain job, the Linux server itself goes unresponsive and no one can even log in to that server.

We constantly get this error message in the logs:
<Free memory in the server is x,xxx,xxx bytes. There is danger of OutOfMemoryError>

What could be the reason and where to start to identify the root cause?

Thanks,



 
Ramakanta Sahoo
Ranch Hand
Posts: 256
Fedora Firefox Browser Netbeans IDE
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
You are having 24managed servers with 4GB Xmxand Xms setting which equals : 24*4 = 96GB

Do you really have that much of memory in system/or combined memory of multiple systems.??

Check how much combined memory is available in each system post 4GB of Xms if not enough memory is available for the system operations then reduce the Xms size to some lower value so that the memory reservation by JVM will be less.
 
sudipto shekhar
Ranch Hand
Posts: 826
Chrome Eclipse IDE Oracle
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
The server logs needs to be checked to identify the exact issue.

You mentioned a certain job, what job is it? What does it do?

In the server logs, do you see any warnings related to any missing performance packages?
 
Sujoy Choudhury
Ranch Hand
Posts: 136
Eclipse IDE Ubuntu
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
@Ramakanta
This is how it was configured and we have high end Linux servers.

@Sudipto
Going forward this is what I am going to do. Just wanted to know before hand if Weblogic cluster was not configured correctly do we generally face these kinds of issues.
But I think, without profiling no one can tell.

Thanks Guys for your valuable comments.
 
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic