40 Posts
0
965
July 30th, 2021 06:00
Can the Networker Server Backup savesets be cloned to multiple pools of different device types?
Any Networker backup system implementation has the default backup policy/workflow ‘Server Protection/Server backup’ to back up the Networker server’s bootstrap and databases. In our Networker system, the server backup savesets are stored in a dedicated backup pool containing a DDBoost device, and cloned to another dedicated clone pool containing a DDBoost device in another remote Data Domain device.
In order to simplify a Networker server recovery in case of a disaster, I would like to keep more server backup saveset copies on two Windows servers, in our case the two storage node servers, one at the local site and another at the remote site. An AFTD device was created on each Windows server and two separated clone pools were created to hold these devices. Two clone actions were added in the workflow ‘Server Protection/Server backup’ to clone the server backup saveset to these new AFTD pools.
When the Networker server needs to be rebuilt, from a new server with empty Networker databases, I can easily copy over the AFTD file folder from one of these Windows servers to the new Networker server; then define a new AFTD device within the new Networker NMC. A scheduled job was also set up on the two Windows storage node servers to run the CLI command ‘mminfo –B’ to record the bootstrap SSIDs. With the bootstrap SSID, it is quite easy to restore the Networker databases by using ‘nsrdr’ command.
I did a simple test with this configuration and it seems working. I am just wondering if having multiple clone actions in a workflow to clone the same saveset to multiple pools with different device types, in our case DDBoost device and ATFD device, is supported or not. In particular, is it a good idea to do this for the ‘Server Protection/Server backup’ workflow? Any feedback is welcome.
bingo.1
2.4K Posts
1
July 30th, 2021 07:00
Of course you can have multiple clone pools to maintain (multiple copies of) your save sets. And of course, the devices/device types do not matter. In general you could copy (clone or stage) from any to any device/media).
The server data is a bit specific as the bootstrap must be sent to a device which is connected locally to the backup server himself. This makes sense as the bootstrap data will rebuild your complex configuration during the DR process. Make it as easy as possible.
BTW it does not matter, whether the DR process will retrieve the data from media of a backup or a backup clone pool.
ChiXC
40 Posts
0
July 30th, 2021 07:00
Thanks bingo.1 for your comments. Yes, the basic idea of this configuration is to avoid configuration of any external device on the new Networker server during the DR process. Assume at least one of the Windows (storage node) servers survived, I can easily copy the AFTD file folder to the new Networker server and get the Bootstrap SSID. Once the Networker server databases are restored, the connections to the external devices, including storage nodes and Data Domain devices, will be automatically restored as well
Djobine_f8450a
5 Posts
0
November 22nd, 2023 16:06
Just to add some details.
I had 2 clone tasks in the server protection workflow (at the same time..., seems not possible to make them concurrent)
The result : clone tasks regularly and randomly failing.
I had to keep only one clone task on the server protection workflow and create a new workflow, with a special group: type Query, include saveserts from the previous 1 hour (I do a bootstrap backup every one hour)
Now this works, I always have 3 copies of my bootstrap/indexes every hours