[
http://jira.jboss.com/jira/browse/JBAS-3659?page=comments#action_12343967 ]
Brian Stansberry commented on JBAS-3659:
----------------------------------------
I would doubt it. This should result in an increase of about 2.3MB in memory used by the
process (not heap memory; this is memory used by the underlying TCP subsystem. Not sure
what the problems you're seeing are; will respond separately to try to trace it down.
Increase ClusterPartition buffers to support farming of large files
-------------------------------------------------------------------
Key: JBAS-3659
URL:
http://jira.jboss.com/jira/browse/JBAS-3659
Project: JBoss Application Server
Issue Type: Task
Security Level: Public(Everyone can see)
Components: Clustering
Reporter: Brian Stansberry
Assigned To: Brian Stansberry
Priority: Minor
Fix For: JBossAS-4.0.5.GA
The receive buffers in the cluster-service.xml UDP protocol are only 150K. Farming sends
files around the cluster in 512K chunks. With the default config, the receive buffer
sizes are not even large enough to hold one chunk. This can lead to problems where the
packets that make up part of a chunk get dropped and have to be retransmitted. With very
large files, this can slow things down to the point that timeouts start getting tripped
and the transfer fails.
We had a support case where increasing the buffers to the sizes used by the
fc-fast-minimalthreads config took care of the problem. That config uses 25MB send
buffers for both mcast and unicast; don't think we need to increase to such a high
number for this. But, I will bump the receive buffers to 640K and the sends to perhaps
2MB.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://jira.jboss.com/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
http://www.atlassian.com/software/jira