HP-UX Cluster Service is not ruuning

[Muhammad Tariq]

when customer executes following command it gives error find below details

cmquerycl -v -C /etc/cmcluster/clusR3P.conf -n dbciR3P

Unable to determine an IP addresses for dbciR3P.

Warning: Unable to determine local domain name for dmgr3prd

Looking for other clusters ... Done

Node dbciR3P is refusing Serviceguard communication.

Please make sure that the proper security access is configured on node

dbciR3P through either file-based access (pre-A.11.16 version) or role-based

access (version A.11.16 or higher) and/or that the host name lookup

on node dbciR3P resolves the IP address correctly.

Failed to gather configuration information

[John Poff]

Hi,

Sounds like a resolution problem. What does this command return?

nslookup dbciR3P

Does it return the correct IP address for that host? If so, try nslookup on that IP address to see that it resolves correctly back to that host.

JP

[Muhammad Tariq]

nslookup working fine with Name and with IP both.

regards.

[David de Beer]

Muhammad,

Did you configure the node access correctly in /etc/cmcluster/cmclnodelist ?

Rgrds,
David.

[Albert_31]

Hello Muhammad,

a) Are you trying to create a single node cluster ?

b) If yes, then please check the /etc/hosts to check how the entries in the file are ?

c) in the name resolution, please check the resolution policy as well

# nslookup
> policy

ensure that you use files first for name resolution.

d) also check for duplicate ip's as well.

netfmt -f /var/adm/nettl.LOGX > /tmp/netlog
X = 00 for 11.00
X = 000 for 11i

post the outputs for us to check.

[Mridul Shrivastava]

check the node names entered in the cluster ascii file. This issue is usually a invalid host/node name or IP address.

1) Check the node name
# uname -n

2) Verify the correct node name and IP address is in the /etc/hosts file.

3) Enter the correct node name and IP into the cluster ascii file as shown below. (example)


NODE_NAME node1 NETWORK_INTERFACE lan0
HEARTBEAT_IP 10.10.10.1
NETWORK_INTERFACE lan1
HEARTBEAT_IP 192.168.193.15
NETWORK_INTERFACE lan2
FIRST_CLUSTER_LOCK_PV /dev/dsk/cxtydz

NODE_NAME node2
NETWORK_INTERFACE lan0
HEARTBEAT_IP 10.10.10.2
NETWORK_INTERFACE lan1
HEARTBEAT_IP 192.168.193.14
NETWORK_INTERFACE lan2
FIRST_CLUSTER_LOCK_PV /dev/dsk/cxtydz


4) Verify the cluster ascii file with cmcheckconf
# cmcheckconf -v -C cluster.ascii

5) Apply the cluster ascii file with cmapplyconf
# cmapplyconf -v -C cluster.ascii

6) Start the cluster with cmruncl
# cmruncl

[Muhammad Tariq]

Hi,
I really appriate yours input, but my hosts file, nodelist, cluster asci ii file everything is OK. I already verfied this, in fact my two node cluster was working fine till thursaday night and it start giving problem from Friday morning even no changes made in the configuration.

regards.
Muhammad Tariq

[Mridul Shrivastava]

could yoy please post the output of the following commands:

#cmviewcl -v
#cmruncl -v

[Matthew Ghofrani]

Hi;

Were you running any packages on this node that is giving you problem? Is the problem constant?

Matthew from Boston

[Steven E. Protter]

Shalom Muhammad,

You must make sure that the hostname dbciR3P resolves correctly and can be pinged from all nodes.

There is a file called cmnodelist that lets you define security, similar to .rhosts

Once all network connectivity issues are resolved, then you can use cmcheckconf and cmapplyconf to rebuild the cluster configuration.

There are no shortcuts with 11.16. You must get no errors from the processes above to make it work.

There are also modifications required to inetd.conf. Please post your configuration and I'll start a system and provide you a working configuration for the two daemons that apply to Serviceguard. Or someone else will.

As an old ITRC user, let me take a few bytes to tell you how it works. If you like the answer you receive and it helps answer your question or solve the problem you posted, it is highly encouraged to assign points to the answers, more points for answers that help more. You may have seen the reminder when you posted to the thread.

Its a really great system and I further encourage you to use your skills to try and help others and earn points.

Regards,

SEP

[Michael Steele_2]

a) Check syslog for major events and commands.

b) Read the service guard logs:

cmreadlog /var/opt/cmom/cmomd.log

cmreadlog /var/opt/sgmgr/929917sgmgr.log

c) ServiceGuard cluster is a binary file. Its possible / probable that this got corrupted. Run :

cmcheckconf -v -C /etc/cmcluster/....ascii -P /etc/cmcluster/.../....conf
-P /etc/cmcluster/...cntl

In lieu of 'cmquerycl' run this command:

cmscancl -n node -o output_file

this may also fail since its an ip issue. but cmscancl provides alot more info for trouble shooting. It will check MAC addresses and connections which is a level lower than IP's.

NoTe: Are you using FQDN? If so don't. serviceguard chokes on FQDN.