cannot use cmviewcl

[zulfi]

Cluster was working fine - I rebooted one fo the nodes and now on this node I get the following message.

Onthe other node - this node status is unknown

Either this node is not configured in a cluster, user doesn't have
access to view the cluster configuration, or there is some obstacle
to viewing the configuration. Check the syslog file for more information.
For a list of possible causes, see the Serviceguard manual for cmviewcl


have amended /etc/rc.config.d/cmcluster to AUTOSTART_CMCLD= ----> 1

[Geoff Wild]

Did you try on the rebooted node:

cmruncl


Next question - why did you reboot?


Next question - did you do a cmhaltnode prior to reboot?

Rgds...Geoff

[melvyn burnard]

on the node you rebooted, do NOT do cmruncl, you must use cmrunnode
Check what it says in the syslog.log, there may be a clue in there.
Also, what version of SG and which patch?
do:
what /usr/lbin/cmcld

[Geoff Wild]

Whoopsie - yes - I meant cmrunnode!

Rgds...Geoff

[zulfi]

Apologies Gentleman got called to a P1 incident.
On the rebooted node I try cmrunnode

Did you try on the rebooted node: - yes and rebooted it again after

cmruncl - cmrunnode and cmviewconf - daemons for cmclconfd is not up and will not come up.

on 2nd host which did not get rebooted HP-ledger:/root# ps -ef | grep cmcl
root 26530 1140 0 16:58:41 ? 0:00 cmclconfd -c -i
root 8279 1140 0 23:15:25 ? 0:00 cmclconfd -p

Also see the cmviewcl on host which did not get rebooted.
HP-ledger:/root# cmviewcl

CLUSTER STATUS
syslog unknown

NODE STATUS STATE
journal unknown unknown
ledger down unknown

UNOWNED_PACKAGES

PACKAGE STATUS STATE AUTO_RUN NODE
loghost down halted enabled unowned

NOTE: can start cluster on this host





Next question - why did you reboot? rebooted as installed syslog-ng - Cluster was set up 6 days ago, was running fine on either node. After installing syslog-ng and it's runtime dependancies. http://hpux.connect.org.uk/hppd/hpux/Networking/Admin/syslog_ng-1.6.11/
Was still unable to get it to work properly, decided to reboot the machine.


Next question - did you do a cmhaltnode prior to reboot? cmhaltpkg then cmhaltcl

Also, what version of SG and which patch?
do:
what /usr/lbin/cmcld

HP-journal:/root# what /usr/lbin/cmcld
/usr/lbin/cmcld:
Cluster Monitor Product $Revision: 82.2 $
Cluster Monitor Product Only $Revision: 82.2 $
Daemon
A.11.18.00 Date: 03/15/07
Build date: Thu Mar 15 16:07:05 PDT 2007
Build id: ibld_sg1118_1123_product
Build platform: hpux

[Geoff Wild]

From the "other node" I assume journal, try:

cmrunnode ledger

Does it start?

If not, ceck the /var/adm/syslog/syslog.log on both hosts.

Rgds...Geoff

[zulfi]

Yes Geoff
runs ok from ledger - see my output from ledger cmviewcl
status for journal is unknown

[zulfi]

HP-ledger:/root# cmviewcl

CLUSTER STATUS
syslog unknown

NODE STATUS STATE
journal unknown unknown
ledger down unknown

UNOWNED_PACKAGES

PACKAGE STATUS STATE AUTO_RUN NODE
loghost down halted enabled unowned
HP-ledger:/root#

[Geoff Wild]

Okay - let me get this straight - you are saying that on journal, if you do a cmrunnode ledger - it works and cmviewcl on journale shows both servers in the cluster?

Yet on ledger, it says unknown?

Or is it unknown on both?

I'm thinking along the lines that from Journal - the node that is working, that you reapply the cluster..

cmcheckconf -C yourcluster.asciifile

no errors? then:
cmapplyconf -C yourcluster.asciifile

Rgds...Geoff

[zulfi]

No Geoff
It's Journal which has the problem - When I cmrunnode on Journal I get
HP-journal:/root# cmrunnode
cmrunnode : Unable to connect to the local cluster: No such file or directory.
Either no cluster configuration file exists, the file is corrupted,
cmclconfd is unable to run, or user root on node journal
doesn't have access to view the configuration.


Yesterday ledger was fine and I could bring the cluster up on it and see my vg's etc.
now on ledger I get
HP-ledger:/sbin/rc2.d# cmrunnode
cmrunnode: Unable to communicate with a running cluster or with all nodes in the cluster.
cmrunnode: In order to use cmrunnode, the cluster must already be running on a subset of reachable nodes or else all cluster nodes must be reachable.

Today I have tried to recreate the ascii file but can;t check it without errors the error on journal after using cquerycl is
Possible Heartbeat IPs:

Warning: No possible heartbeat networks found.
Not probing node journal as it is currently unreachable.

All nodes must be connected to at least one common network.
Cannot connect to configuration daemon (cmclconfd) on node journal

[zulfi]

Thanks gentleman
Stephen Doud solved - inetd was not running on the server after the reboot.

this did the trick

[zulfi]

thank you again gentlemen

Problem resolved with inetd being started - weird how it didn't come up on reboot

Dropped you some points for your efforts.