-
1. Re: JBoss 5.1.0.GA CLuster configuration
wdfink Dec 8, 2011 1:55 AM (in response to camiloarfv)The default configuration is without cluster.
You should use the all configuration!
-
2. Re: JBoss 5.1.0.GA CLuster configuration
camiloarfv Dec 9, 2011 7:53 AM (in response to wdfink)the "default" configuration is the configuration that brings jboss is obvious that if I set c luster is "all", ie I have the initial configuration of "all" in jboss for the cluster.
-
3. Re: JBoss 5.1.0.GA CLuster configuration
wdfink Dec 9, 2011 8:05 AM (in response to camiloarfv)Does your JBoss now clustered or do you have still problems?
I suppose you are on different phys. boxes because of the IP and for that I suppose you might be problems with multicast.
Mcast should work in the same subnet without any problem.
-
4. Re: JBoss 5.1.0.GA CLuster configuration
camiloarfv Dec 9, 2011 8:11 AM (in response to wdfink)what you're saying is that I have two jboss on the same subnet or will not serve themulticast configuration.
because I need is to connect to a machine that is on a different subnet, in this casecould start as jboss?
-
5. Re: JBoss 5.1.0.GA CLuster configuration
wdfink Dec 9, 2011 9:52 AM (in response to camiloarfv)You should check whether multicast is possible with [1].
If it is possible you should see messages like 'Number of members: 2' in your logfile if you start two servers.
If not you might ask the network admin to configure multicast between the two subnetworks,
or you change to tcp instead of multicast with '
run.sh -Djboss.default.jgroups.stack=tcp
...'
and activate tcp ping as discovery protocol see [2,3]
[1] http://community.jboss.org/wiki/TestingJBoss
-
6. Re: JBoss 5.1.0.GA CLuster configuration
camiloarfv Dec 9, 2011 11:05 AM (in response to wdfink)Thank you very much!
the cluster is good but in the final of log show this:
2011-12-09 10:35:47,539 WARN [org.jgroups.protocols.pbcast.GMS] (ViewHandler,prueba,172.20.0.73:7800) 172.20.0.73:7800 failed to collect all ACKs (1) for unicasted view [172.20.0.73:7800|2] [172.20.0.73:7800, 172.20.0.113:7801, 197.0.2.186:7802] after 5000ms, missing ACKs from [197.0.2.186:7802] (received=[]), local_addr=172.20.0.73:7800
2011-12-09 10:35:48,695 WARN [org.jgroups.protocols.pbcast.GMS] (ViewHandler,prueba-HAPartitionCache,172.20.0.73:7800) 172.20.0.73:7800 failed to collect all ACKs (1) for unicasted view [172.20.0.73:7800|2] [172.20.0.73:7800, 172.20.0.113:7801, 197.0.2.186:7802] after 5000ms, missing ACKs from [197.0.2.186:7802] (received=[]), local_addr=172.20.0.73:7800
2011-12-09 10:35:48,711 WARN [org.jgroups.protocols.FD_SOCK] (OOB-7,172.20.0.73:7800) I was suspected by 197.0.2.186:7802; ignoring the SUSPECT message
2011-12-09 10:35:50,883 WARN [org.jgroups.blocks.ConnectionTable] (ConnectionTable.AcceptorThread,172.20.0.73:7801) Could not read accept connection from peer java.net.SocketException: Connection reset
2011-12-09 10:35:51,430 WARN [org.jgroups.blocks.ConnectionTable] (ConnectionTable.AcceptorThread,172.20.0.73:7801) Could not read accept connection from peer java.net.SocketException: Connection reset
2011-12-09 10:36:19,665 ERROR [org.jgroups.protocols.pbcast.NAKACK] (Incoming-2,172.20.0.73:7801) sender 172.20.0.113:7900 not found in xmit_table
2011-12-09 10:36:19,665 ERROR [org.jgroups.protocols.pbcast.NAKACK] (Incoming-2,172.20.0.73:7801) range is null
-
7. Re: JBoss 5.1.0.GA CLuster configuration
wdfink Dec 10, 2011 3:20 PM (in response to camiloarfv)Is it only once or periodical.
Looks like that one node is suspected and drop from the cluster but will rejoin.