Actions Performed by the Geographic Edition Framework During a Switchover

Language:

A switchover consists of the following actions:

Application services are unmanaged on the former primary cluster.
The replication role is reversed and now continues to run from the new primary.
Application services are brought online on the new primary cluster.

The geopg switchover command performs the following actions:

The command confirms that the secondary cluster does allow the replication role reversal.
The command creates copies of the replicated source projects and exports the shares for use on the target appliance, to ensure that no projects and mount point conflicts can occur.
The command destroys the created clones on the target appliance and confirms that the actual reverse replication operation can be performed on the target appliance.

Note - When migrating EMC SRDF replication services, basic Geographic Edition operations such as geopg switchover perform a symrdf swap operation. The symrdf swap operation requires significantly more time for static RDF than dynamic RDF. Therefore, you might need to increase the value of the Timeout property of the protection group when using static RDF.

The following data replication products have additional behaviors:

Hitachi TrueCopy and Universal Replicator:

The replication subsystem checks that the Hitachi TrueCopy or Universal Replicator device group is in a valid aggregate device group state.
The replication subsystem checks that the local device group states on the target primary cluster, cluster-newyork, are 23, 33, 43, or 53.

The local device group state is returned by the pairvolchk -g device-group-name -ss command. These values correspond to a PVOL_PAIR or SVOL_PAIR state.

The Hitachi TrueCopy or Universal Replicator commands that are run on the new primary cluster, cluster-newyork, are described in the following table.

Aggregate Device Group State	Valid Device Group State on Local Cluster	Hitachi TrueCopy and Universal Replicator Switchover Commands That Are Run on `cluster-newyork`
SMPL	None	None
Regular primary	`23`, `43`	No command is run, because the Hitachi TrueCopy or Universal Replicator device group is already in the `PVOL_PAIR` state.
Regular secondary	`33`, `53`	`horctakeover –g dg [–t]` The`–t` option is specified when the `fence_level` of the Hitachi TrueCopy or Universal Replicator device group is `async`. The value is calculated as 80% of the `Timeout` property of the protection group. For example, if the protection group has a `Timeout` value of 200 seconds, the value of `–t` used in this command is 80% of 200 seconds, or 160 seconds.
Takeover primary	None	None
Takeover secondary	None	None

Oracle Data Guard:
1. Before the switchover, the command checks that the remote database is in an enabled state in the Oracle Data Guard Broker configuration. The command also confirms that the configuration is healthy by issuing the Oracle Data Guard command-line interface (dgmgrl) show configuration command to ensure that the command returns a SUCCESS state.
2. If the output from this command indicates that Oracle Data Guard Broker is busy performing its own health check, the Oracle Data Guard command-line interface retries the command until it receives a SUCCESS response or until two minutes have passed. If the command-line interface is unable to get a SUCCESS response, the command fails.
Oracle ZFS Storage Appliance:
1. The command creates copies of the replicated source projects and exports the shares for use on the target appliance to ensure that no projects and mount point conflicts can occur.
2. The command destroys such created clones on the target appliance and confirms that the actual reverse replication operation can be performed on the target appliance.

The command performs the following actions on the original primary cluster:

For Availability Suite, removes affinities and resource dependencies between all the application resource groups in the protection group and the internal resource group, such as the lightweight resource groups
Takes offline the application resource groups in the protection group and places them in the Unmanaged state.
Performs a switchover to the secondary cluster for each replication configuration in the protection group.

On the original secondary cluster, the command takes the following actions:

Runs the script that is defined in the RoleChange_ActionCmd property
Brings online all application resource groups in the protection group

If the command completes successfully, the secondary cluster becomes the new primary cluster for the protection group. The original primary cluster becomes the new secondary cluster. The application resource groups in the protection group are brought online on the new primary cluster and replication from the appliance that is connected from the new primary cluster to the new secondary cluster begins.

Note - For Hitachi TrueCopy or Universal Replicator, after a successful switchover, at the data replication level the roles of the primary and secondary volumes have been switched. The PVOL_PAIR volumes that were in place before the switchover become the SVOL_PAIR volumes. The SVOL_PAIR volumes in place before the switchover become the PVOL_PAIR volumes. Data replication will continue from the new PVOL_PAIR volumes to the new SVOL_PAIR volumes.

The Local-role property of the protection group is also switched regardless of whether the application could be brought online on the new primary cluster as part of the switchover operation. On the cluster on which the protection group had a Local role of Secondary, the Local-role property of the protection group becomes Primary. On the cluster on which the protection group had a Local-role of Primary, the Local-role property of the protection group becomes Secondary.

Note - For Oracle Data Guard, databases that are associated with the Oracle Data Guard Broker configurations of the protection group have their role reversed according to the role of the protection group on the local cluster. For HA for Oracle configurations, the dataguard_role resource property is also updated with the status of the new primary and standby clusters. The shadow Oracle database server resource group and any other application resource groups in the protection group are online on the new primary cluster. Data replication from the new primary cluster to the new standby cluster begins.