Engine instance takes long time to iInitialize in a Multi-Instance engine system
Article ID: KB0092699
Updated On:
| Products | Versions |
|---|---|
| TIBCO DataSynapse GridServer | - |
| Not Applicable | - |
Description
Resolution:
The main cause of the problem is that on Linux and Solaris, java.security.SecureRandom().netLong() takes a long time to complete if there are multiple processes trying to access /dev/random.
More details of the symptoms can be found in: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6202721
This happens when launching multiple Engine invoke processes on Linux and Solaris machines. Since multiple invoke processes are blocked in random seed initialization, some invoke processes timeout before the JVM is initialized and the daemon will restart the invoke process. This creates even longer delay in Engine start time.
The workaround is to set -Djava.security.egd=file:/dev/./urandom in Engine JVM Command-line Arguments in the Engine configuration.
The main cause of the problem is that on Linux and Solaris, java.security.SecureRandom().netLong() takes a long time to complete if there are multiple processes trying to access /dev/random.
More details of the symptoms can be found in: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6202721
This happens when launching multiple Engine invoke processes on Linux and Solaris machines. Since multiple invoke processes are blocked in random seed initialization, some invoke processes timeout before the JVM is initialized and the daemon will restart the invoke process. This creates even longer delay in Engine start time.
The workaround is to set -Djava.security.egd=file:/dev/./urandom in Engine JVM Command-line Arguments in the Engine configuration.
Issue/Introduction
Engine instance takes long time to iInitialize in a Multi-Instance engine system
Feedback
Yes
No
Powered by
