On 7 Aug 2009, at 11:02, Galder Zamarreno wrote:
Mircea, what are the following system properties used for?
MYTESTIPS=10.16.93.249 10.16.93.250,
MYTESTIP_2=10.16.93.250,
MYTESTIP_1=10.16.93.249,
MCAST_ADDR=227.43.88.186
MCAST_PORT=45874
Are these valid IPs in the machine where it's been run?
Maybe that MACT_ADDR and MCAST_PORT is being used by hudson for
something and messes up things? Doubt this is the problem but
wondering.
When you see that it fails, do u see any WARN or ERROR messages in
the logs? I do remember seeing stuff like this when the testsuite
wouldn't work on ATL cluster:
infinispan.log:201:2009-07-16 04:30:27,051 WARN
[org.jgroups.protocols.TCP] (pool-3-thread-1) failed to join /
224.0.75.75:7500 on eth2: java.net.SocketException: No such device
I know this is the diagnostics address but when tests run fine, I
never saw this warn message.
On 08/06/2009 11:16 PM, Mircea Markus wrote:
> Hi,
>
> I've looked into the test suite failures on hudson and they are
> mainly
> caused by the fact that cluster is not formed.
> I've switched back from TCP+MPING to TCP+PING and that solved the
> problem (there are only 3 errors in core module now, still have to
> look
> into that).
BTW I believe that there was a problem with using TCP+PING with
JGroups 2.8, and listing all of the initial_hosts. Possibly something
to do with the logical addresses, that wasn't an issue with JGroups 2.6.
http://lists.jboss.org/pipermail/infinispan-dev/2009-June/000718.html
http://lists.jboss.org/pipermail/infinispan-dev/2009-June/000719.html
Cheers
Manik
--
Manik Surtani
manik(a)jboss.org
Lead, Infinispan
Lead, JBoss Cache
http://www.infinispan.org
http://www.jbosscache.org