cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted

Query regarding one vcs node down

Hi Team,

Please suggest, if I have 4 nodes and due to some network issue suddenly all 4 nodes down. 3 nodes came up but 1 node is not coming up.

So what would happen to vcs services.Does that run automatically or need to run it manually.

Please share the step by step activity.

Thanks.

Allaboutunix

1 Solution

Accepted Solutions
Highlighted
Accepted Solution!

Re: Query regarding one vcs node down

Now we can start from scratch. What is the OS on your cluster nodes? If Unix/Linux, check contents of /etc/gabtab It should contain something like this : /sbin/gabconfig -c -n4 This means that vcs will only start with all 4 nodes seeded. Best not to change this number by manually seeding the nodes that are up: gabconfig -cx Then issue hastart. See this TN : http://www.veritas.com/docs/000041575 Best to read up about split brain and ways to prevent that with sufficient network redundancy and I/O fencing.

View solution in original post

7 Replies
Highlighted

Re: Query regarding one vcs node down

Please provide me the solution.

 

Thanks..

Highlighted

Re: Query regarding one vcs node down

You will get a quicker and more appropriate response by posting to the appropriate product specific forum.  This is a forum for general community discussion and forum related issues.

Highlighted

Re: Query regarding one vcs node down

Could you please tell  me in which forum I have to write the query. This is for veritas cluster service product.

 

Thanks..

Highlighted

Re: Query regarding one vcs node down

Have move to (what I hope is) the correct forum for you.

Highlighted

Re: Query regarding one vcs node down

Ok Andy.. many Thanks.. to provide the way..

 

Thanks..

Highlighted

Re: Query regarding one vcs node down

Hi Team,

 

Please suggest, waiting for your response.

Thanks..

Highlighted
Accepted Solution!

Re: Query regarding one vcs node down

Now we can start from scratch. What is the OS on your cluster nodes? If Unix/Linux, check contents of /etc/gabtab It should contain something like this : /sbin/gabconfig -c -n4 This means that vcs will only start with all 4 nodes seeded. Best not to change this number by manually seeding the nodes that are up: gabconfig -cx Then issue hastart. See this TN : http://www.veritas.com/docs/000041575 Best to read up about split brain and ways to prevent that with sufficient network redundancy and I/O fencing.

View solution in original post