• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 2757
  • Last Modified:

weblogic memory

Folks

Can i increase my memory settings on weblogic for each managed server, currently they are:

set MEM_ARGS=-Xms512m -Xmx1536m -XX:MaxPermSize=512m

i am using 64 bit weblogic 923 with JROCKIT  R27.5.0-jdk1.5.0_14

pls advise


r
0
rutgermons
Asked:
rutgermons
  • 3
  • 2
1 Solution
 
ECollinCommented:
i should say "why not ?"
-First of all, in a production environment, you should set -Xms at the same value as -Xmx : -Xms:1536m -Xmx:1536m
-Are you facing memory issues ?
-What kind of applications are deployed in your domain : Web app, portal, ejbs... ?

Emmanuel
0
 
rutgermonsAuthor Commented:
Hi Emmanuel

We deploy maximo which is a web app via weblogic  (maximo.ear) file


we had maxpermsize memory issues at 256 but increased it to 512 and it seems better, our big issue seems if we run complex queries from the front end on a cluster then the users would get kicked out of the application

the funny thing is, is that using a standalone appserver pointing to the same database we cannot replicate this error, this leads me to think the iss plugin is problematic, I also see a lot of multicast messages are getting lost and i get an error message in wlproxy.log:

Thu Sep 09 01:58:14 2010 <1764128401519417> failure on sendRequest() w/ recycled connection to 149.177.182.132:8080, numfailures=1, lastError=0
Thu Sep 09 01:58:14 2010 <1764128401519417> Marking 149.177.182.132:8080 as bad
Thu Sep 09 01:58:14 2010 <1764128401519417> Exception occurred for backend host '149.177.182.132/8080/0' while sending request : 'PROTOCOL_ERROR [line 842 of ..\nsapi\.\URL.cpp]: Backend Server not responding'
Thu Sep 09 01:58:14 2010 <1764128401519417> got exception in sendRequest phase: PROTOCOL_ERROR [line 842 of ..\nsapi\.\URL.cpp]: Backend Server not responding at line 959; last_error 0


it seems our managed servers are losing connection to the cluster thats why the kickouts

any ideas?

0
 
ECollinCommented:
you can try to replace your IIS by a simple web app deployed in a stand alone weblogic server that will act as a proxy in front of your cluster (via HttpClusterServlet).  That's for checking your cluster communications.
Related doc : http://download.oracle.com/docs/cd/E13222_01/wls/docs92/cluster/setup.html#wp684345
0
 
rutgermonsAuthor Commented:
will it handle the big user load? we have 200 concurrent users over 14 managed servers?
0
 
ECollinCommented:
woah !
So many servers for so few concurrent users...
Well tuned i think you could handle all your users with only one managed server...
How many concurent http sessions do you have to handle ?
Why do you need a cluster ?
Emmanuel
0

Featured Post

Free Tool: Site Down Detector

Helpful to verify reports of your own downtime, or to double check a downed website you are trying to access.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

  • 3
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now