Restore the MongoDB data on an alternate cluster
NetBackup supports the following alternate recovery scenarios for MongoDB:
Redirected restore and recovery to an alternate cluster
Redirected restore and recovery to an alternate node or port or database path in an existing cluster
To restore the MongoDB data on an alternate cluster
- Run the tpconfig command to update the original cluster's credentials with the alternate application server's credentials.
For example, to recover source client
Host1-26050to an alternate application serverHost2that is running on port 28001:Add the credentials of
Host2:28001and its related nodes in the original cluster's credential configuration file.Run the update tpconfig command for application_server that you want to recover (Host1-26050)
Here is a sample command:
/usr/openv/volmgr/bin/tpconfig -update -application_server Host1-26050 -application_type mongodb -requiredport 26050 -application_server_conf /usr/openv/var/global/credential.conf
- Open the NetBackup web UI.
- On the left, click Recovery.
- On the Regular recovery card, click Start recovery.
- In Basic properties, enter the following:
Select the Policy type as BigData > MongoDB
From the Source client list, select the required Application server.
From the Destination client list, select the required backup host. Restore is faster if the backup host is the media server that had backed up the node.
Click Next.
- In Recovery details, do the following:
If necessary, click Edit to select the appropriate date range to restore the complete data set. Or select Use backup history and select the backup images that you want to restore.
Select the files and folders for restore.
Click Next.
- In Recovery options, do the following:
Rename the application server and its nodes and set the value for the alternate application server.
Select Restore individual directories and files to different locations. To change the folder paths, click Edit file paths.
In the MongoDB options, select Restore and recover and recover databases for current time or select the specific schedule.
Click Next.
- Review and click Start recovery.
You can check the status in the Activity monitor.
You can restore the MongoDB oplog files from an incremental backup to an alternate path. The files and their path are seen in the web UI.
You must specify the paths during the alternate restore using the option.
If you want to retain the original MongoDB path but change the oplog file path, click and specify the source and alternate paths.
For example, Source /host:port/tmp and Destination /host:port/alternate_tmp.
For an alternate restore from a nested database path, use the Add Destination dialog box and for every subfolder, add an appropriate target alternate path.
For example, to change the path from /host:port/usr/mongodb/db1 to /host:port/alt-dir/dbpath/mydb:
Specify the source and the destination path:
Source /host:port/usr/mongodb/db1 and Destination /host:port/alt-dir/dbpath/mydb
Specify the source and the destination path for the parent folder:
Source /host:port/usr/mongodb and Destination /host:port/alt-dir/dbpath
Specify the source and the destination path for the base parent folder:
Source /host:port/usr and Destination /host:port/alt-dir
Note:
When you do an alternate restore to a non-root path, the restore is partially successful if the database path contains multiple subfolders.
In such a scenario, when you do an alternate restore to a different location, you must add an entry for each directory level.
For example:
Source:/hostname1:port1/Config_Data
Destination: /hostname2:port3/mongo_inst2
Source:/hostname1:port1/Config_Data/data
Destination:/hostname2:port3/mongo_inst2/data
Source:/hostname2:port2/Shard1_Primary
Destination:/hostname2:port3/mongo_inst2
Source:/hostname2:port2/Shard1_Primary/data
Destination:/hostname2:port3/mongo_inst2/data
You can restore a MongoDB cluster (sharded or replica set) that was backed up from different nodes. This capability exists because of the role switch (between primary and secondary nodes) within a shard or a replica set. In this scenario, the full backup can be taken from one host and the incremental backup is taken from another host in the same shard or replica set.
During restore, you must redirect the restore of these backup images to the same MongoDB host.
For example, to restore backups from /host1:port1/dbpath and /host2:port1/tmp, specify the following:
Source /host1:port1/dbpath and Destination /althost:port1/dbpath
Source /host2:port1/tmp and Destination /althost:port1/tmp