Service Gurad not starting after upgrading the patch and kernel

[Ganesan R]

Hi All,

We have two node cluster running on RHEL4. Recently we upgraded the patch level including kernel on one node and afterwards cluster services are not starting on patched node.

This is the error we are getting while try to start the cluster services.

# cmrunnode -v node2
Failed to load data from configuration database.
cmrunnode: Unable to determine nodes on the current cluster.
Either no cluster configuration file exists, the file is corrupted,
cmclconfd is unable to run, or user root on node node2
doesn't have access to view the configuration.

Upon investigation I found that "identd" daemon is not running on this node. It is not getting started also.

#/sbin/init.d/identd start
FATAL: Module pidentd not found.

Is this cluster service not starting due to identd daemon not running on this node?

How to I fix this error and start identd daemon ?

Hope you people can give the solution ?

[Matti_Kurkela]

You forgot to recompile the "deadman" and "pidentd" modules, which are critical parts of Serviceguard for Linux. These must be recompiled for each kernel upgrade.

As you're now running the upgraded kernel, the procedure is simple. First make sure you have the kernel-smp-devel (or kernel-devel, if you use single-processor servers) package upgraded to match exactly the version of your kernel package.

Then, for deadman:

cd /usr/local/cmcluster/drivers
make clean
make
make install

 And for pidentd:

cd /usr/src/pidentd-3.0.15sg/drivers
make clean
make
make install

 When both modules are successfully compiled, run "depmod -a", then reboot.

Identd and Serviceguard should now start normally.