RedHat Cluster Suite - No Quorum
 

Hi,

I have a running RedHat cluster with two nodes. On both of the nodes the clustat utility says:

Cluster Status - httpd_cluster 14:27:47
Incarnation #1
(This member is not part of the cluster quorum)

Member Status
------------------ ----------
wwwb Active
wwwa Active <-- You are here

No Quorum - Service States Unknown


Does somebody know why?

The manual is very poor unfortunately, it says nothing about this message: (This member is not part of the cluster quorum)

Where can I more info on RedHat Cluster Suite?

vamosb

I encounter the same problem . Fortunately,I have solved it. Format the sharedisk on each node. Until each node can read same file on sharedisk. Then setup the Quorum , it worked. I think this solution is not exactly correct:),just solve the problem.

Thanks.

What does 'Setup the Quorum' mean for you?

set th Quorum in clustermanager

What partition should I format? On the shared disk there are two partition for two raw device. I think I do not need to format them.

You should format the raw device as ext3,and make a file for read by each node to ensure share disk worked.

Solved
 

There was a network problem between the two nodes. After it was resolved, the quorum started.

That's good news.

I have a similar problem, in this case we cant see the nodes active at the same time:

Clustat second node:
Cluster Status - cluster 10:26:58
Incarnation #1
(This member is not part of the cluster quorum)

Member Status
------------------ ----------
172.16.0.49 Active <-- You are here
172.16.0.48 Inactive

No Quorum - Service States Unknown

clustat first node
Cluster Status - cluster 10:26:42
Cluster Quorum Incarnation #1
Shared State: Shared Raw Device Driver v1.2

Member Status
------------------ ----------
172.16.0.49 Inactive
172.16.0.48 Active <-- You are here

Service Status Owner (Last) Last Transition Chk Restarts
-------------- -------- ---------------- --------------- --- --------

second node:
[root]# service clumanager status
Se está ejecutando clumembd (pid 2673)...
Se está ejecutando cluquorumd (pid 2671)...
Se está ejecutando clulockd (pid 2678)...
clusvcmgrd está parado
Note: Service manager is not running because this member
is not participating in the cluster quorum.

first node
[root]# service clumanager status
Se está ejecutando clumembd (pid 5547)...
Se está ejecutando cluquorumd (pid 5545)...
Se está ejecutando clulockd (pid 5552)...
Se está ejecutando clusvcmgrd (pid 5577)...

However if a run:

#/sbin/shutil -p /cluster/header

I view the same information in the both nodes

Thanks for your advice

The RedHat cluster suite needs more than 50% of members to start quorum.
If you use Disk Tiebreaker, this resouce give one more point to cluster, but is not recommended to cluster with 2 members.
If you have a network problem, you need to give one more point to cluster using "cluforce member_name"
see: man cluforce