Help with Cluster SetUp

[Joe]

Hi forum friends.

As part of a future clustering implemetation in my work,
i'm developing a simulated 2 cluster nodes using VMWARE
(before do the real job, of course). I have the W2K White
Papers, etc.

My question is: how can i setup the 2 nodes without a
real/shared storage? All documents and papers point that a
physical share storage is 100% needed in order to get the
cluster up and running (i know that). But... could be a
workaround for this and setup the cluster without a
physical disk array? just VMWARE?...

Any help/suggestions will be very appreciated,

Thanks

 

[John Toner [MVP]]

Joe,

Shared storage is required for a working cluster.

If you want to simulate shared storage in VMWARE, have a look at the
following web page for some very helpful tips:

http://www.vmware.com/pdf/ESXCluster.pdf

Regards,
John

 

[Joe]

Thanks for reply.

I found this link too and i would like to share with you:

http://techrepublic.com.com/5100-6268_11-5034801.html

I was able to follow each steps till the cluster set up.
When finish, i recieved an error indicating that cluster
service could not start. And since then... i can't give
any step ahead...i don't know why the cluster service
can't start

 

[John Toner [MVP]]

What is the error?

Thanks for reply.

I found this link too and i would like to share with you:

http://techrepublic.com.com/5100-6268_11-5034801.html

I was able to follow each steps till the cluster set up.
When finish, i recieved an error indicating that cluster
service could not start. And since then... i can't give
any step ahead...i don't know why the cluster service
can't start

 

[Joe]

Thanks for reply John.

This is the error:
There are no more endpoints available from the endpoint
mapper.

I must say that i'm able to see/ping/resolve machines
between 'em.

 

[John Toner [MVP]]

That's a generic RPC error.

A couple questions...Did you try a reboot? Is the RPC service
enabled/started? Did you service pack the cluster nodes? Are the nodes
members of a domain?

If this doesn't help, send the last 20 or so lines of the cluster.log so we
can see if there's any hint in there.

Regards,
John