Hello Mr. Nouri,

My name's Carmine Moleti, and i'm writing you because i've been
struggling with the
very same issue as you...

I have managed to get rid of the "I/O error on channel" and make the
o2cb resource
start on the cluster!

Going through the strace output on "ocfs2_controld.pcmk" i have
spotted the place
where the program initiate the handshake with the kernel OCFS2 module (as you
did).

Then i went reading the relevant code and found that the somewhat
cryptic message
was due the fact that the local node id was far bigger than the allowed maximum
for the kernel code!!!

Thus an ERANGE return code is returned from the handshaking code which triggers
the cryptic message.

Enough words! :)

The step to solve the issue is as simple as adding the following line
to corosync.conf file,
under the totem secion:

clear_node_high_bit: yes

it's also possible to specify a fixed node id!

Hope this helps.

Cheers,
Carmine
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to