Re: Re: [Q] How does Oracle 9i-RAC handle lock to suport HA?

> Now, it's not really that simple, because the GCS CAN'T be on one node
> only - single point of failure which is what we want to get away from for
> HA. The GCS is actually duplicated on each node that is running an Oracle
> instance.

First of all, thanks for your answer.

Let's take a simple example of a 3 node cluster with instances 1, 2 and 3 (located on nodes 1, 2 and 3).
Suppose instance1 is the master of resource group 1, instance 2 is the master of resource group 2 and instance 3 is the master of resource group 3. Does "The GCS is actually duplicated on each node ~~~" mean that the GCS and GES information about resources managed by instance 1 is duplicated on instance 1 and 2(for example)?

One example,

 resources               master              managing information
                                                       is duplicated by
-----------------------------------------------------
resource group 1     Instance 1          instance 1, 2
resource group 2     Instance 2          instance 2, 3
resource group 3     Instance 3          instance 3, 1
-----------------------------------------------------