Once you have installed and configured your multicluster
primary host, add one or more
multicluster failover
hosts to your multicluster
primary cluster. Setting up
multicluster
primary failover is optional,
but is recommended for production clusters. Should your primary host fail, failover hosts
take over.
About this task
This process involves installing the multicluster package on at least
one other host to be your failover host, completing configuration steps on that host, and defining a
primary candidates list to
indicate the host names for your primary and failover hosts.
Procedure
-
Install the multicluster package on another
host (that is, one that is not your multicluster
primary host; this new host
will be your multicluster
failover host).
-
Configure your multicluster failover host.
Follow the same steps as Configuring the multicluster primary host; however,
if you are configuring on Linux, you do not need to set
entitlement on additional multicluster hosts; you only need
to set entitlement once on the multicluster
primary host.
-
Repeat steps 1 and 2 to install and configure on additional multicluster failover hosts.
-
Configure the primary
candidates list. This is the list of hosts that take over should the multicluster
primary host fail.
Run the following command from any one of your
multicluster failover hosts (note
that regardless of the number of failover hosts you have installed and configured, you only need to
run this command once from any one of the failover hosts):
Tip: The
following steps use the egoconfig administrative command. Ensure that you are
logged on as cluster administrator to run
egoconfig.
egoconfig masterlist primary_host,failover_host_1,failover_host_2
where:
- primary_host is the host name of your multicluster
primary host. Always list the
primary host first, when you
run this command.
- failover_host_number is the host name of your multicluster failover hosts. If
you have installed and configured more than one failover host, lists them all here, separating each
by a comma. List them in the order of preference for failover. Always follow the primary host with a list of failover
hosts, when you run this command.
Note: This procedure applies to in-data center failover, not to cross-data center
recovery. This means that the failover host is only for a primary host in the same cluster.
Cross-data center recovery involves two separate clusters.
For example, if your primary
host is hostP and you have one failover host, hostX, run:
egoconfig masterlist hostP,hostX
What to do next
Once you have set up failover hosts to your multicluster
primary cluster,
you can start the multicluster
primary, and then test your
entire multicluster
primary cluster.