cmclquery Error: Unable to establish communication to node mcsgnode1: 19

[Thomas Stuetzer]

Hello,
I want to setup a MC/SG cluster for testing purpose. So my first step is to create a cluster configuration ASCII file.
But I get an error
sapr3man:/#cmquerycl -v -n mcsgnode2 -n mcsgnode1 -C /etc/cmcluster/cmclconf.ascii


Begin checking the nodes...
Warning: Unable to determine local domain name for sapr3man
Looking for other clusters ... Done
Error: Unable to establish communication to node mcsgnode1: 19
Failed to gather configuration information.


in the syslog appears this message on both nodes:
sapr3man:/#tail /var/adm/syslog/syslog.log
Jun 22 16:13:32 sapr3man cmclconfd[19301]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory

Any idea?
thanks Thomas


Please find hereunder some config details

sapr3man:/etc/cmcluster#nslookup mcsgnode1
Using /etc/hosts on: sapr3man

looking up FILES
Name: sapr3man
Address: 138.200.248.22
Aliases: sapr3manb, sapr3man.aat.de, mcsgnode1

sapr3man:/etc/cmcluster#nslookup mcsgnode2
Using /etc/hosts on: sapr3man

looking up FILES
Name: sandbb
Address: 138.200.248.12
Aliases: mcsgnode2

sapr3man:/#more .rhosts
mcsgnode2 root
mcsgnode1 root


sapr3man:/grep ha /etc/inetd.conf
hacl-probe stream tcp nowait root /opt/cmom/lbin/cmomd /opt/cmom/lbin/cmomd -i -f /var/opt/cmom/cmomd.log -r /var/opt/cmom
hacl-cfg dgram udp wait root /usr/lbin/cmclconfd cmclconfd -p
hacl-cfg stream tcp nowait root /usr/lbin/cmclconfd cmclconfd -c -i

sapr3man:/#netstat -a |grep ha
tcp 0 0 *.chargen *.* LISTEN
tcp 0 0 *.hacl-probe *.* LISTEN
tcp 0 0 *.hacl-cfg *.* LISTEN
udp 0 0 *.chargen *.*
udp 0 0 *.hacl-cfg *.*

sapr3man:/#more /etc/nsswitch.conf
services: files[NOTFOUND=continue] nis
hosts: files
aliases: files

sapr3man:/#uname -a
HP-UX sapr3man B.11.11 U 9000/800 542746541 unlimited-user license


sapr3man:/#swlist |grep -i service
B3935DA A.11.15.00 MC / ServiceGuard
B5140BA A.11.11.03 MC/ServiceGuard NFS Toolkit
B7609BA A.04.00.01 Event Monitoring Service
B7885BA B.03.09 MC/Service Guard Extension for SAP

The second Cluster node has the same config.

[Mustafa Gulercan]

hi;
edit following files for both nodes;
/etc/hosts
/.rhosts
/etc/cmcluster/cmclnodelist

and check /etc/inetd.conf;
there shouldn't be any # before

ident stream tcp wait bin /usr/lbin/identd identd



regards;
mustafa

[Sundar_7]

Can you try adding the all the 3 host aliases to the .rhosts file (I prefer /etc/cmcluster/cmclnodelist though)

# vi ~root/.rhosts
sapr3man root
sapr3manb root
sapr3man.aat.de root
mcsgnode1 root
mcsgnode2 root
#

[Thomas Stuetzer]

Hello Mustafa and Sundar,

I have checked the files /etc/hosts ~root/.rhosts /etc/cmcluster/cmclnodelist
and /etc/inetd.conf and haven't see any misspelling. I just have added in the aliases to those files, but the error is still exist.
Thomas

[Mustafa Gulercan]

try this command please;

node1> cmquerycl -v -k -C /etc/cmcluster/cmclconf.ascii -n node 1 -n node2

regards;
mustafa

[David de Beer · Accepted solution]

Thomas,

Is your /etc/resolv.conf correct? I see that you are not using DNS, but perhaps the two files on the machine are inconsistent.

If you can, use the latest version of MCSG - if you've linked your support contract to your profile you will be able to download the latest version of MCSG for 11i? I assume you are using 11i? Use 11.16 - 11.17 is available, but I have only seen it for 11.v2.

Are you going to use Oracle RAC on these clusters?

With 11.16 you can use /etc/cmcluster/cmclnodelist, which imho opinion is better and more secure than using .rhosts

Also, is there any reason you are using aliases for the hostnames? We've always just used the hostnames for creating clusters.

Also, instead of doing a cmquerycl (which is used to get the config of EXISTING clusters), use cmcheckconf ;-)

The documentation for 11.16 is here -> http://docs.hp.com/en/B3936-90079/index.html

The documentation for 11.15 is here -> http://docs.hp.com/en/B3935-90063/index.html

Kind regards,
David.

[Thomas Stuetzer]

Hello David and Mustafa,

I have deleted the aliases in the file /etc/hosts. So that hostname , systemname and network resolve (nslookup)
is pointing to the same name.
After that I had botted the server.
Now the command cmquerycl is working fine.

Thomas