Unsolved
This post is more than 5 years old
4 Posts
0
828
October 6th, 2014 07:00
Recover returns: cannot support multiple clients
I'm trying to recover a directory with the following command on a Windows box:
recover -s server -d c:\temp -S 1042943833
where server has a DD640 mounted as a CIFS volume.
The error is same as another post that was attributed to incorrectly spelled name. This is not the case:
recover: cannot support multiple clients, ssid 3516032769 from client 'hostname' and not 'null'
client hostname is the hosts whose data I am trying to recover.
Background. I have Networker server/client installed at my home institution. It has a DD620 mounted as CIFS volume. This is where our backups go. This DD620 is replicated replicated to our parent institution's DD640. I have a second Networker sever/client installed there. I was able to scan the replicated volume and mount it. What I was trying to do is recover ANY hosts data based on an SSID from my remote location, testing disaster recovery.
The error message above, could it be due to the fact that the Networker server that put the data there is not the same node that is trying to recover it?
Any help would be appreciated.
0 events found
crazyrov
4 Operator
•
1.3K Posts
1
October 6th, 2014 08:00
You will have to create the client instance at the destination server with the same client id as that of the source.
crazyrov
4 Operator
•
1.3K Posts
0
October 6th, 2014 09:00
Incase you have a separate NetWorker zone in the destination with its own set of machines being backed up then yes you will have to. Else you could do a DR of the Networker server at your destination assuming you have the bootstrap information replicated over as well.
ble1
4 Operator
•
14.4K Posts
0
October 6th, 2014 09:00
Given that you are starting DR with backup server, you should use same data as on primary site. If you replicate those, then mdb which contains clientid and res which contains resources will be in place already. I do realize that in this case this is not classic DR, but as part of your DR readiness you must have procedure which keeps necessary data available. In your case, I would first test the theory and create client as it is on original server using same clientid and test same restore again. That will test the theory. If this works, then you just need to enhance your steps by having something like mminfo -avot -q client,clientid | sort -u output being ready somewhere.
klnasveschuk
4 Posts
1
October 6th, 2014 09:00
So in event of a disaster, I would need to know all the client IDs to be recover any data? What if all I had was the replicated data. Anyway to get around that?