If it becomes necessary to delete Managed Server information under the circumstances specified below, this section provides a way to delete the Managed Server information forcibly from the Management Server.
The operating system on the Managed Server cannot be started because of a hardware failure or another similar reason. From that point, the server is not used for operations.
The Managed Server has been physically removed without performing the Managed Server deletion procedure ("3.6.4 Deleting Managed Server").
The Managed Server is managed by AdvancedCopy Manager and has been physically removed without being deleted in AdvancedCopy Manager.
For cases other than these, refer to "3.6.4 Deleting Managed Server" to delete the Managed Server.
Note
When this procedure is performed, the specified Managed Server is forcibly deleted regardless of the backup or replication status. Therefore, this procedure should be used only for Managed Server that are not planned to be used afterwards.
Do not execute a backup or replication while this procedure is performed.
The components and management information of AdvancedCopy Manager on the forcibly deleted Managed Server are not deleted.
When the forcibly deleted server is to be used again, uninstall the AdvancedCopy Manager from the deleted server before it is reconnected to the network. If its server is reconnected to the networks without uninstalling first AdvancedCopy Manager, the AdvancedCopy Manager functions of other servers on the network may not work correctly.
The procedure to delete forcibly a Managed Server is described below:
Confirm that the Managed Server is not used afterwards or that it is removed before starting the procedure.
Check the status of AdvancedCopy Manager on the Managed Server to be deleted forcibly.
If AdvancedCopy Manager is operable, use the normal deletion procedure for deleting the Managed Server instead of the forcible deletion procedure. For details about the normal deletion procedure, refer to "3.6.4 Deleting Managed Server".
On the Management Server, execute the stgxfwcmsetdev command with -n option. Specify the target Managed Server with an operand of -n option.
After executing the command, check the status of Managed Server as follows.
Go to step 2, only when the Managed Server cannot be started.
When the Managed Server cannot be started:
The following message is displayed:
"swstf5405 The acquisition of device information caused a failure.
Reason=swstf2111 Connection to Storage Server failed."
When the Managed Server can be started:
The following message is displayed:
"swstf5408 There is no change in information currently being managed."
Alternatively the command may be waiting for input of the device information confirmation. If the command is waiting for input, enter 'q' or 'n' to prevent updating the device information and end the stgxfwcmsetdev command execution.
Stop the Advanced Copy sessions.
Even if the Managed Server information has been deleted, the Advanced Copy sessions for the connected disks (Logical Units) may not be stopped.
For the unnecessary Advanced Copy sessions, stop its sessions by using ETERNUS Web GUI.
Back up the database.
For the procedures, refer to "8.7.2.1 Backup of Common Control and AdvancedCopy Manager Repository".
However, for the "pg_dump" command, read the commands below.
# /opt/FJSVesfcm/postgres/bin/pg_dump -U esfpostgres -a --attribute-inserts -p portNumber -t esccs.\"BASE_\"* -t esccs.\"REP_REPVOL\" -t esccs.\"BR_\"* -f $BAK_DIR/Common/fileName esfdb
Delete the Managed Server information forcibly.
Execute the stgforcedelsrv command.
Note
If the stgforcedelsrv command terminated abnormally, do not proceed from step 5 onwards.
If the steps after step 5 are executed an incorrect database is restored and normal system operation may not be performed adequately afterwards.
Restore the database.
Restore the database from the directory that the database is stored. Its directory must be a directory of "after-dir" specified with -a option in the stgforcedelsrv command at step 4.
Prepare for the restoration.
For non-cluster operations, execute the following shell script to stop the service.
# /opt/FJSVesfcm/bin/stop-webservice.sh
For cluster operations, perform the following procedure:
Stop all cluster transactions which the Manager belongs to.
Refer to "Cluster Transaction and Local Transaction" in the Cluster Environment Setup Guide for details about cluster transaction.
Mount the shared disks on the primary node.
Start the ETERNUS SF Manager Postgres service.
# /etc/init.d/startSFmanager start-postgres
Restore a backup data. The following procedure must be implemented by a user with administrator privileges for the operating system.
Execute the following command to restore the backup data.
If you did not change the port number, specify 15432 for portNumber. If you did change the port number, specify the changed port number for portNumber.
The fileName is the file name specified when the backup data was created.
# /opt/FJSVesfcm/postgres/bin/psql -U esfpostgres -p portNumber -f $BAK_DIR/Common/fileName esfdb
The prompt is displayed, and waiting for input. Input the following strings.
master01!
Start the Manager.
For non-cluster operations, execute the following shell script to start the service.
# /opt/FJSVesfcm/bin/start-webservice.sh
For cluster operations, perform the following procedure:
Stop the ETERNUS SF Manager Postgres service.
# /etc/init.d/startSFmanager stop-postgres
Unmount the shared disk mounted in step a.
Start all cluster transactions which the Manager belongs to.
Perform the following operation according to the status of the forcibly deleted Managed Server.
If a replication operation has been done:
If the forcibly deleted Managed Server was part of a server-to-server replication:
On the Management Server, execute the swsrprecoverres command with the -h and -r options. For the operand of -h option, specify the name of the server used for the replication with the forcibly deleted Managed Server. If several servers where used for replication, the swsrprecoverres command has to be executed for each of those servers.
If the forcibly deleted Managed Server was not part of a server-to-server replication:
No operation is required.
If a backup operation has been done:
No operation is required.
Take the following actions to reflect to Web Console depending on how the agent installed in the forcibly deleted Managed Server is used:
If both of AdvancedCopy Manager's agent and Storage Cruiser's agent are used:
With Web Console, execute the Reload Conf. operation to update to the latest information.
If only AdvancedCopy Manager's agent is used:
With Web Console, delete the Managed Server that you have deleted.
