Service Guard: can't start identd daemon

[Kraft, Wolfram]

I'm presentlx configuring HP Service Guard on Linux (SLES9). Some SG cmds return messages like: #####
cmviewcl : Cannot find cluster cluster1.
Either cluster cluster1 is not configured, or the user doesn't
have access to view the cluster configuration. #####
My assumption is that cause is the not-running ident daemon. Starting identd fails:
#######
in.identd[18491]: started
in.identd[18491]: bind(5,[::]:113) failed: Address already in use
in.identd[18491]: terminating
#######
anyboby any suggestion ?

regards

Wolfram

[Peter Godron]

Wolfram,
please check your inetd.conf and services files.
Do you have an auth and an ident line. If try to run with the auth line commented out.
Can you check what is running on the port specified in services (netstat or lsof)?

If the answer was useful, please see:
http://forums1.itrc.hp.com/service/forums/helptips.do?#28

Willkommen zum ITRC Forum!

[Serviceguard for Linux]

Wolfram,
Did you uninstall authd?

rpm -e authd

[Kraft, Wolfram]

Hi all,

answer 1: did not solve my problem

answer 2: how can I identify the package ? is authd part of the pam-suite and can I de-install the complete pam-suite ?

regards

Wolfram

[Serviceguard for Linux]

The documentation details how to de-install authd.

It may not have been clear but the line I had list was the command you need to execute:

# rpm -e authd

This is in the release notes.

Regards

[Kraft, Wolfram]

I'm afraid, but "rpm -e authd" returns:

recworus140:/opt/cmcluster/conf # rpm -e authd
error: package authd is not installed
recworus140:

You hopefully understand my confusion.

regards

Wolfram

[Serviceguard for Linux]

Sorry - I missed the fact that you are running SLES9 when I saw your identd comment. There are a lot of possible issues that can cause messages similar to the one you have.

I assume that the entire cluster configuration process went OK. When you executed "runcl" were there any messages in the system log?

Are you executing the cmviewcl command on one of the cluster nodes and are you root at that time? I ask that because of access control policies. I'm not sure if that is the error if you do not have the access control pocies set up correctly, but it may be what you see.

If you think that identd is not running (and I think you can check yast for that) then try cmapplyconf again. If identd is not running, I'm pretty sure this should fail.

(I'm home & do not have access to a cluster right now to try this.)

Regards

[Kraft, Wolfram]

Hi,

i'm not that far with the configuration. I just could do a cmquerycl -v -C .... - L /dev/... -n node1 -n node2

Trying a cmviewcl -v -c ... results in:

#############
recworus140:/oracle/ECP/oraarch # cmviewcl -v -c cluster1
cmviewcl : Cannot find cluster cluster1.
Either cluster cluster1 is not configured, or the user doesn't
have access to view the cluster configuration.
##############

installed rpm's:

nfs-toolkit-A.01.03-0.product.suse.i386.rpm
pidentd-3.1a27sg-1.x86_64.rpm
serviceguard-A.11.16.07-0.product.suse.x86_64.rpm
sgcmom-B.03.01.02-0.product.suse.x86_64.rpm
sgesap-toolkit-A.02.00-0.product.suse.x86_64.rpm

[Kraft, Wolfram]

sorry,

problem has been solved long ago, just forgot to close it