Solved

heap rises and idle threads go to 0

Posted on 2006-11-09
7
659 Views
Last Modified: 2013-12-10
I have a problem with a weblogic express cluster that i run.

It is a weblogic 8.1.4 3 node cluster running on RHES4.

every 2 or so days the heap size rises to 95% and then the idle threads drops to 0. then the sites become unavlable.

Any ideas where to start looking?
0
Comment
Question by:willa666
7 Comments
 
LVL 5

Expert Comment

by:Becky
ID: 17910600
It could be garbage collection gone bad...  

Do you have a dev server you can test on?  The easiest thing I can think of is to start your app on a test machine, then about 1 and 1/2 days later manually run garbage collection and see if it exhibits the same behavior while GC is running.
0
 
LVL 1

Author Comment

by:willa666
ID: 17927173
Hi  greensunie

If it is a issue with garbage collection then how can it be fixed?

Will
0
 
LVL 5

Expert Comment

by:Becky
ID: 17928211
You'd need to get a really experienced coder to go over your applications on that server and look out for places where memory leaks can be caused.   Not closing database connections properly (and not using a finally clause) are a major cause of memory leaks (and can cause the GC to run early and often).  Other types of connections to other computers such as LDAP servers, other kind of authentication mechanisms, any kind of socke programming, etc.. can all cause leaks.
0
The Eight Noble Truths of Backup and Recovery

How can IT departments tackle the challenges of a Big Data world? This white paper provides a roadmap to success and helps companies ensure that all their data is safe and secure, no matter if it resides on-premise with physical or virtual machines or in the cloud.

 
LVL 5

Expert Comment

by:Becky
ID: 17928212
socke = socket
0
 
LVL 6

Accepted Solution

by:
Ajar earned 500 total points
ID: 17931239
Hi willa,

    There seems to be a particular servlet/service on your website that might need cure. In order to identify this , the next time your server becomes non responsive ,open the admin console and monitor the execute queue for the non responsive server. [DOMAIN>servers>server>monitor all execute queues]It probably will turn out that most of the  threads are servicing a similiar request.  Once you have identified this request you know the part of application that is mis behaving.

    You can further get the THREAD DUMP of the server threads by  pressing Ctrl + Break on windows  and  Kill -3 <wls pid> on unix. This will get you a fair idea about the classes that are keeping the threads engaged.




0
 
LVL 4

Expert Comment

by:illusionz70
ID: 17963590
HI willa ,
Are you using any middleware ? like Tuxedo/JOLT or something like that between you app server and database.
If yes Check for the connection pool not getting released.
0
 
LVL 23

Expert Comment

by:rama_krishna580
ID: 18046831
0

Featured Post

Master Your Team's Linux and Cloud Stack

Come see why top tech companies like Mailchimp and Media Temple use Linux Academy to build their employee training programs.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Suggested Solutions

-Xmx and -Xms are the two JVM options often used to tune JVM heap size.   Here are some common mistakes made when using them:   Assume BigApp is a java class file for the below examples. 1.         Missing m, M, g or G at the end …
This exercise is about for the following scenario: Dmgr and One node with 2 application server. Each application server contains it owns application. Application server name as follows server1 contains app1 server2 contains app1 Prereq…
Nobody understands Phishing better than an anti-spam company. That’s why we are providing Phishing Awareness Training to our customers. According to a report by Verizon, only 3% of targeted users report malicious emails to management. With compan…

831 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question