Server startup takes a long time on a federated node
 Technote (troubleshooting)
 
This document applies only to the following language version(s):
English
 
Problem(Abstract)
On a federated node, if nodeagent or one of the servers is already started, then the next server takes more than seven minutes to start.

<services xmi:type="loggingservice.ras:RASLoggingService" xmi:id="RASLoggingService_1" enable="true" messageFilterLevel="NONE"
enableCorrelationId="true" suppressStackTraces="false">
<serviceLog xmi:id="ServiceLog_1" name="${LOG_ROOT}/activity.log"
size="10" enabled="true"/>
</services>
 
Cause
If Activity.log size is different for each server in server.xml, then second JVM will try to change the size causing this problem.

Turn trace on with com.ibm.ejs.ras.*=all=enabled specification and start the server to verify that following trace entries observed.


[4/27/05 6:23:04:824 UTC] 65b75e04 SharedLogWrit > SharedLogWriter - ctor
[4/27/05 6:23:04:836 UTC] 65b75e04 SharedLogWrit > verifyLogFile
[4/27/05 6:23:04:838 UTC] 65b75e04 SharedLogWrit > processExistingFile
[4/27/05 6:23:04:839 UTC] 65b75e04 SharedLogWrit d processExistingFile - changing size of log file
[4/27/05 6:23:04:958 UTC] 65b75e04 SharedLogRead e Ctor the free space size of 2084561 did not match the physical file size 10485760
[4/27/05 6:23:04:958 UTC] 65b75e04 SharedLogRead d fillBuffer - reading 3eb bytes from 52 .... lot more of above
[4/27/05 6:29:41:709 UTC] 65b75e04 SharedLogRead d fillBuffer - firstHalf reading 28d bytes from location 9ffd73
[4/27/05 6:29:41:709 UTC] 65b75e04 SharedLogRead d fillBuffer - 2nd read 15e bytes from location 52
[4/27/05 6:29:41:709 UTC] 65b75e04 SharedLogRead d checkForMessage checking location 52
[4/27/05 6:29:41:709 UTC] 65b75e04 SharedLogRead d fillBuffer - reading 4 bytes from 52 [4/27/05 6:29:41:709 UTC] 65b75e04 SharedLogRead > readBlockSize reading blockSize from location 56
[4/27/05 6:29:41:709 UTC] 65b75e04 SharedLogRead d fillBuffer - reading 4 bytes from 56 [4/27/05 6:29:41:709 UTC] 65b75e04 SharedLogRead < readBlockSize returning blockSize = 14b
[4/27/05 6:29:41:710 UTC] 65b75e04 SharedLogRead d fillBuffer - reading 4 bytes from 1a5

 
Resolving the problem
Change the Activity.log size to be the same for all servers and nodeagent on the federated node.

Use Administrative Console to modify Activity.log size. Verify that Maximum File Size is same for all JVMs on this node.

Servers -> Application Servers -> <server name> -> Logging and Tracing -> IBM Service Logs >> Maximum File Size

 
 
Cross Reference information
Segment Product Component Platform Version Edition
Application Servers Runtimes for Java Technology Java SDK
 
 


Document Information


Product categories: Software > Application Servers > Distributed Application & Web Servers > WebSphere Application Server > Hangs/Performance Degradation
Operating system(s): AIX
Software version: 5.1.1.3
Software edition:
Reference #: 1205252
IBM Group: Software Group
Modified date: Apr 30, 2005