Clustering - EhCache Replication

[Eduardo P García, geändert vor 12 Jahren.]

Hi there,

We have displayed a clustered Liferay 6.0.4 CE installation with ehcache set in multi-vm-clustered mode. When we run performance tests (i.e. load simulation with Jmeter), we observe that:

1. There is too much traffic between the members of the cluster. We guess this is due to replication, but CPU is hight everytime. It seems replication never stops, despite we make no changes in the portal... We are using default RMI replication.

2. Having a look at Ehcache Beans status using VisualVM, we find that some beans (especially com.liferay.portal.velocity.LiferayResourceCacheUtil) have too many cache misses.

3. Adding extra nodes actually seems to decrease the overall performance of the system.

Anyone with experience in such issues could please provide some hints? Clustering documentation does not cover much troubleshooting.

Thank you very much in advance!

[Michal Wroblewski, geändert vor 11 Jahren.]

Hi,

did you find answer to your problem ?

Regards,
M

[J B, geändert vor 11 Jahren.]

attach your portal-ext.properties

[Michael Stratford, geändert vor 11 Jahren.]

Hi

i have similar problem (i think).
Below my post:
http://www.liferay.com/community/forums/-/message_boards/message/22219902

Mayby you can help ?

Thanks in advance
Michael

[Jack Chung, geändert vor 11 Jahren.]

This may not work for liferay 6.0.4 CE, but did work for Liferay 6.1.10. Note that all the files in the ehcache directory in the portal-impl.jar has to be extracted out to the myehcache directory. Have not tested performance fully, but did not find any usual issues with high CPU usage. This does use jgroups instead of RMI so that account for some of the performance differences.

portal-ext.properties:

net.sf.ehcache.configurationResourceName=/myehcache/hibernate-clustered.xml

net.sf.ehcache.configurationResourceName.peerProviderProperties=peerDiscovery=automatic,multicastGroupAddress=${multicast.group.address["hibernate"]},multicastGroupPort=${multicast.group.port["hibernate"]},timeToLive=1

ehcache.multi.vm.config.location=/myehcache/liferay-multi-vm-clustered.xml

ehcache.multi.vm.config.location.peerProviderProperties=peerDiscovery=automatic,multicastGroupAddress=${multicast.group.address["multi-vm"]},multicastGroupPort=${multicast.group.port["multi-vm"]},timeToLive=1 

ehcache.bootstrap.cache.loader.factory=com.liferay.portal.cache.ehcache.JGroupsBootstrapCacheLoaderFactory

ehcache.cache.event.listener.factory=net.sf.ehcache.distribution.jgroups.JGroupsCacheReplicatorFactory

ehcache.cache.manager.peer.provider.factory=net.sf.ehcache.distribution.jgroups.JGroupsCacheManagerPeerProviderFactory

[Gunta Dalecka, geändert vor 10 Jahren.]

So, have you succeeded to configure Liferay cluster with working session replication?
Which Liferay version did you use? I try to do it with Liferay 6.1.2 CE.
How did you do it? I tried both with and without Terracota.
Did you use session.shared.attributes in your solution? I have troubles with it, when configuring cluster with Terracota.
I've asked related questions in another forum post. Any ideas, how to solve them?