Unalbe to create node in Redhat clustering

fosiul01
fosiul01 used Ask the Experts™
on
Hi,
I am trying to create a node from luci server. its actually placing cluster.conf file in the node call node2.domain.local
but problem is , in cluster.conf file its missing the node name. hence , i am seeing error in /var/log/messages like this

Sep 15 13:33:05 node2 ccsd[5332]: Starting ccsd 2.0.98:
Sep 15 13:33:05 node2 ccsd[5332]:  Built: Aug 17 2009 15:01:41
Sep 15 13:33:05 node2 ccsd[5332]:  Copyright (C) Red Hat, Inc.  2004  All rights reserved.
Sep 15 13:33:05 node2 ccsd[5332]: cluster.conf (cluster name = test-cluster, version = 1) found.
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] AIS Executive Service RELEASE 'subrev 1358 version 0.80.3'
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] Copyright (C) 2002-2006 MontaVista Software, Inc and contributors.
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] Copyright (C) 2006 Red Hat, Inc.
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] AIS Executive Service: started and ready to provide service.
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] local node name "node2.xxx.local" not found in cluster.conf
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] Error reading CCS info, cannot start
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] Error reading config from CCS
Sep 15 13:33:07 node2 openais[5338]: [MAIN ] AIS Executive exiting (reason: could not read the main configuration file).
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] AIS Executive Service RELEASE 'subrev 1358 version 0.80.3'
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] Copyright (C) 2002-2006 MontaVista Software, Inc and contributors.
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] Copyright (C) 2006 Red Hat, Inc.
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] AIS Executive Service: started and ready to provide service.
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] local node name "node2.xxx.local" not found in cluster.conf
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] Error reading CCS info, cannot start
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] Error reading config from CCS
Sep 15 13:33:08 node2 openais[5364]: [MAIN ] AIS Executive exiting (reason: could not read the main configuration file).
Sep 15 13:33:34 node2 ccsd[5332]: Unable to connect to cluster infrastructure after 30 seconds.

i am stuck now
i am looking at google from last more then 3 hours
its looks like /etc/hosts file
i have changed /etc/hosts file so many way
currently its liket his

193.132.234.240  node2.elect.local node2

but no luck

i also tryed wiht this /etc/hosts

127.0.0.1 node2.electro.local
193.132.234.240  node2.elect.local node2

if i try to start service cman start
it will say
cman not started : cant find a local node name in cluster.conf /usr/sbin/cman_tool:aisezec daemon did not start

as i said, node name ismissing from cluster.conf, but why ???

how shall i fixed the issue??


note : I am trying to do a 2 node clustering
Comment
Watch Question

Do more with

Expert Office
EXPERT OFFICE® is a registered trademark of EXPERTS EXCHANGE®
cjl7freelance for hire
Commented:
Hmm,

You should really read the document I sent you in the previous post, there is a howto on what you need to do before you start the installation.

http://www.redhat.com/docs/manuals/csgfs/

//jonas
Top Expert 2009

Author

Commented:
hi ur and mine time difference is 24 hrs i guess! It 12 am here. Anyway. I tryed to create 2nd node whole day but no luck. Every pc has giving me same error. Making luci server and make the luci server as 1 cluster is fine. But adding another pc as 2nd node doesnt. Acordinng to those document ,from luci server, i can add another node which i m doing,i can see luci server installing software in 2nd node. But problem seems like it dont write cluster node name in cluster.conf file in 2nd node server. Hence those error. Really stuck. Which part of those link actually saying .how to create 2nd node? And how to solve those issue. Really appreciate ur suggestions
Top Expert 2009

Author

Commented:
or do i need to create 2nd node via systrm-config-cluster software from master node.?
Introduction to R

R is considered the predominant language for data scientist and statisticians. Learn how to use R for your own data science projects.

Top Expert 2009

Author

Commented:
hi
today i went through those article again and again, but no luck
i reinstall the linux os to the node again, then i started again from luci server by adding a new node. and same, its createing cluster.conf file in node2.mydomain.local
but it would not put nodename in their!
so i copyed file from main cluster to node2.mydomain.local

after that, its just showing

Sep 15 13:33:34 node2 ccsd[5332]: Unable to connect to cluster infrastructure after 30 seconds.

i tryed the bellow articles and files, but still now luck


https://bugzilla.redhat.com/show_bug.cgi?id=217724
http://sources.redhat.com/cluster/wiki/FAQ/CMAN
https://bugzilla.redhat.com/show_bug.cgi?id=213946

all of them saying its due to /etc/hosts file


my /etc/hosts file is like this


# Do not remove the following line, or various programs
# that require network functionality will fail.
127.0.0.1               localhost.localdomain localhost
::1             localhost6.localdomain6 localhost6
193.132.234.58 node2 node2.mydomain.local


but no luck...

do you have any idea whats i am doing wrong ??

can you little bit specifiq please
Top Expert 2009

Author

Commented:
just extra add to my previous post

on node2, cman is not running
when i am trying to run cman service

i get these error:

Failed
/usr/sbin/cman_tool: aisexec daemon didnot start

and those link i provided before was for this error

please some one give me some light
cjl7freelance for hire
Commented:
you need to put all the hosts in the hosts file.

<ip addr> <name of 1st node>
<ip addr> <name of 2nd node>
<ip addr> <name of 3rd node>

and so on.

Also make sure that you ssh to all the hosts on every possible address, ip-address, short-name, fqdn, localhost.

Then run ricci on all hosts and luci on one of them.

//jonas
Top Expert 2009

Author

Commented:
HI
I just did what you said

its a 2 hosts cluster
node1
node2

in /etc/hosts file, i added both host file this way


In Node 1 :

127.0.0.1           localhost
193.132.234.6   beaver beaver.internal.local
193.132.234.58 node2 node2.internal.local


in node2: ( problematic one)

193.132.234.6   beaver beaver.internal.local
193.132.234.58 node2 node2.internal.local localhost


Also : i copyed cluster.conf from node1 to node2 ( is that Ok ??)

Now when i restart cman in node2
it will tell you

Failed
/usr/sbin/cman_tool: aisexec daemon didnot start

and in the log file

Sep 15 13:33:34 node2 ccsd[5332]: Unable to connect to cluster infrastructure after 30 seconds.



i dont know what to do ......

Top Expert 2009

Author

Commented:
and this is the cluster.conf file

<?xml version="1.0"?>
<cluster alias="test-cluster" config_version="31" name="test-cluster">
      <fence_daemon clean_start="0" post_fail_delay="0" post_join_delay="3"/>
      <clusternodes>
            <clusternode name="beaver.electroversal.local" nodeid="1" votes="1">
                  <fence/>
            </clusternode>
            <clusternode name="node2.electroversal.local" nodeid="2" votes="1">
                  <fence/>
            </clusternode>
      </clusternodes>
      <cman expected_votes="1" two_node="1"/>
      <fencedevices/>
      <rm>
            <failoverdomains/>
            <resources/>
      </rm>
</cluster>




NOte: this file mainly created by node1( where is luci server) then i copyed it over node2
Top Expert 2009

Author

Commented:
hi ya

have a look to this question

http://www.experts-exchange.com/OS/Linux/Q_24739039.html#a25357438


is there any known issue for cman version
cman-2.0.115  ??
Top Expert 2009

Author

Commented:
read this one

https://www.centos.org/modules/newbb/viewtopic.php?post_id=85661&topic_id=22250

i will be reinstall both system again with older version of cman
Top Expert 2009
Commented:
yap thats the problem!!

cman-2.0.115  does not work!!

i installed cman-2.0.98-1.el5 in both node, and its working fine

and yes, cman-2.0.115  has some issue, i dont know yet how to solved but i will look into this later on

thanks for the help

Do more with

Expert Office
Submit tech questions to Ask the Experts™ at any time to receive solutions, advice, and new ideas from leading industry professionals.

Start 7-Day Free Trial