On 02/22/2011 08:02 AM, [email protected] wrote:
Hi there,
I'm trying to build qpid 0.8 with support of clustering on the RedHat
Enterprise Linux 6.
For the clustering support to be available it is necessary corosync
headers and object files to be available, but in there are only
corosync_1.2.3-21.el6.x86_64.rpm, corosynclib_1.2.3-21.el6.x86_64.rpm,
corosync_1.2.3-21.el6.i686.rpm.
You need to install corosynclib-devel for the header files.
I have tried to build corosync-1.2.3, and use its headers and object files
to build qpid and I was able to do it, but now I have another problem.
When I'm trying to restart corosync and qpid on reboot, process of
corosync dies as well as process of the qpid.
The last error message in the log of the corosync (from the RHEL
distribution) is
Feb 22 15:58:30 corosync [CPG ] lib_init_fn: conn=0x201dc90,
cpd=0x201c370
Here is log of the qpid
2011-02-22 15:58:30 notice Journal "TplStore": Created
2011-02-22 15:58:30 notice Store module initialized;
store-dir=/opt/qpid-0.8/data/cluster-node1
2011-02-22 15:58:30 notice Initializing CPG
2011-02-22 15:58:36 error Error during CPG shutdown: Cannot finalize CPG
group : library (2)
2011-02-22 15:58:36 critical Unexpected error: Cannot get local CPG
identity: library (2)
2011-02-22 15:58:36 critical Unexpected error: Daemon startup failed:
Cannot get local CPG identity: library (2)
When corosync and qpid are started as services manually there are no any
problems.
Could you please suggest how to troubleshoot this problem? It seems that
such a problem occurs only on system reboot.
Try the checklist at https://cwiki.apache.org/qpid/starting-a-cluster.html
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:[email protected]
