Just like hp_brs said make a backup of /etc/cmcluster, then cmdeleteconf.,
and cmapplyconf. On the node being removed a few cleanups,
comment out hacl's in /etc/inetd.conf and /etc/services. (execute inetd -c
to re-read conf. ) This will drop the ports that SG
listen on and communicates over. Then remove the node from the
/etc/cmcluster/cmclnode list and/or /.rhosts. Cleanup /etc/lvmrc so the vg's
can auto activate again on boot.
That should do it , good luck.

"BB" <ccc@caraudio.com> wrote in message
news:Pc25d.146042$Q7D.289@twister01.bloor.is.net.c able.rogers.com...
> We have just take a node out of cluster (permanently). Cluster config and
> pkg files have been carefully modified. cmcheckconf comes out clean with
no
> syntax errs. Now, if I do cmapplyconf will this cause any sorts of
problems
> at all? Will the cluster know that node is nolonger in service right away?
> if a node does go down will this out of commission node still try to
> "adopt"?
>
> Has anyone saw a glitch? Its just if there's a change of cluster becoming
> unstable I don't want to do it live. Hence, looking for
advise/opinions/past
> experience!
>
> ENV:SG A.11.15 on 11.11 (N4000)
>
> Thank you all.
>
>