[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-04-15 Thread karnivas
Any thoughts on this ??? Please reply. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4225552#4225552 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4225552 ___ jboss-user maili

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-04-14 Thread karnivas
Just to give an update on this, it was identified that the dead member's(suspected member's) neighbor was having a wrong gateway entry in the routing table configuration. Instead of having its own IP it was pointing to another IP which was also in the same cluster. Once that instance was rest

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-04-02 Thread karnivas
We brought down one of the SOLARIS machine(P1 - Co-ordinator) to check the view in all machines. As expected, the co-ordinator changed to one of the RHEL machine by removing the P1 from all views, but the dead RHEL members wasn't updated in the VIEW Please find the DEBUG messages of jgroups.log

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-30 Thread bstansbe...@jboss.com
First, I recommend you try the tests discussed at http://www.jboss.org/community/docs/DOC-12375[/url] and [url]http://www.jgroups.org/manual/html/ch02.html#ItDoesntWork. There could be a communication problem between the RHEL nodes and the Solaris nodes. If all is well there, recommend you tur

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-30 Thread karnivas
"karnivas" wrote : We tested the same with JBoss Cache 1.4.1 SP8 with JGroups 2.4.1.SP8 but the problem still exists We tested the same with JBoss Cache 1.4.1 SP8 with JGroups 2.4.1.SP4 but the problem still exists View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-30 Thread karnivas
We tested the same with JBoss Cache 1.4.1 SP8 with JGroups 2.4.1.SP8 but the problem still exists View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4221971#4221971 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4221971

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-27 Thread karnivas
One of the RHEL instances has detected the failure and has the following messages org.jgroups.protocols.FD --> [SUSPECT] suspect hdr is [FD: SUSPECT (suspected_mbrs=[xx.xx.xx.xx:xx)] org.jgroups.protocols.VERIFY_SUSPECT --> verifying that xx.xx.xx.xx is dead But this message is not updated to t

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-26 Thread karnivas
YES. The viewAccepted in the SOLARIS machines still have the dead members after some hours also. How can these members be deleted from the VIEW ? Should we set any other value to resolve this ?? View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4221456#4221456 R

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-26 Thread bstansbe...@jboss.com
With the config the old RHEL server instance should be removed from the view after 7.5 seconds (6 secs for FD to suspect it, plus 1.5 for VERIFY_SUSPECT). Are you saying you are still seeing a view that has two incarnations of the RHEL server in it? Is this after at least 7.5 seconds have passed

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-26 Thread karnivas
When one of the RHEL instance was brought down by killing the instance forcefully (kill -9 {PID}), the VIEW in the SOLARIS machines aren't updated even after testing with your suggestion. The sample config is as follows | | | |

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-26 Thread bstansbe...@jboss.com
Use both FD and FD_SOCK. See explanation at http://www.jboss.org/community/docs/DOC-9739 View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4221372#4221372 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4221372 _

[jboss-user] [JBoss Cache: Core Edition] - Re: FD Issue

2009-03-26 Thread karnivas
The added code was the sample so the IP(bind_addr) is different but all the other configuration is the same. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4221195#4221195 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4