cmmquerry is giving error in A11.19 SG

[Sreer]

Hi Gurus,

While doing the cmquerryscl for frming a cluster iam getting the following errors:

I hav 2 nodes:
Node1
Node2


# cmquerycl -v -C /etc/cmcluster/cluster.conf -n Node1 -n Node2 -q quorum
Warning: Unable to determine local domain name for Node1
Looking for other clusters ... Done
Node Node2 is refusing Serviceguard communication.
Please make sure that the proper security access is configured on node
Node2 through either file-based authorization(pre-A.11.16 version) or
access control policies (version A.11.16 or higher)
and/or that the host name lookup on node Node2
resolves the IP address correctly.
Failed to gather configuration information.


thank you very much for yiour advice.

OS => 11.13

cmclnodelist is perfect


rgds Sree

[SoorajCleris]

Hi,

What the system says when you say

#nslookup hostname ( as mentioned in cmclconfig)

what is there in your nsswitch.conf

Regards,
Sooraj

Please paste the output of #cat cmclnodelist

[Vijaykumar_1]

Whether the /etc/hosts file on both the nodes identical ?

Do you have any network connectivity issues betwee nthe nodes ..

[rariasn]

HI,

cluster.conf file:

NODE_NAME is the specified nodename in the cluster.

It must match the hostname and both cannot contain full domain name.

rgs,

[melvyn burnard]

it appears you may have not set the entries in the /etyc/cmcluster/cmclnodelist file correctly.
Refer to the Securing Serviceguard manual and the Managing Serviceguard manuals on the document web site

[Vijaykumar_1]

The entry should be as follows,

nodename/ip username #node1
nodename/ip username #node2


This should reside in /etc/cmcluster directory.

[Stephen Doud]

There are several causes of 'refusing Serviceguard communication':

1) Insure all three hacl lines are uncommented in /etc/inetd.conf, 16 in /etc/services.

Insure the 'auth' line is uncommented in /etc/inetd.conf and has a partner line in /etc/services. If you modify the file, run 'inetd -c' to make inetd re-read the file.

2) Every fixed IP listed in "netstat -in" output must be aliased to the nodename on which it is assigned in /etc/hosts. (It is required per the Managing Serviceguard manual).

3) cmclnodelist must be configured with both hostnames/IPs for root on BOTH nodes.

4) If the server was Ignited from an image from another cluster node, remove the invalid cluster binary (/etc/cmcluster/cmclconfig)

5) Use the real hostnames in cmquerycl, including uppercase if any.

6) With the cluster down, kill old cmclconfd -p processes. If it keeps restarting, something keeps running SG commands, which can hold off cmquerycl etc.

7) If /var/adm/inetd.sec is configured to specifically 'allow' Serviceguard services, those lines must also include the loopback (127.0.0.1) IP.
Example:
auth allow 10.* 129.71.* 172.16.* 169.254.* 127.0.0.1

[Sreer]

Thanks all !

I have used only primary lan and the heartbeat also just to start the cluster.
And now it started.

I will use it once the cluster start functioning.

I dont know y it happend. Any way worke d after making this change in NIC configuration.

Thanks all for your support.
I assigned poits as usual for your extreme support :--)

Thnks Rgds Sree