Unable to backup Server 2

[logicacmg000]

Ever since we change the server's IP address we are unable to back it up. We are using LegatoNW 6.1.4 on Unix and trying to backup a Win2k3 server. I changed the IP address on the Win servers's host file and also the Backup server's host file. But I get this error.

--- Unsuccessful Save Sets ---

* "servername":All nsrexec: authtype nsrexec
* "servername":All savefs: unknown host name: "backup server-name (as in /etc/hosts)"

Any clues?
Thanks.

 

[new2me]

How about your DNS server?

Can you ping the server_name, etc.

 

[logicacmg000]

Yeah I am able to ping the client from the server and vice-versa, using the new IP address and the servername.

Thanks.

 

[605]

Although you can not find any docs about it, i think the IP address is stored with the client-id which is generated when the client is defined and saved in the media index. I think this is where the problem is.

First, try whether you can run a manual backup at all.

Or do the following:
- Shut down NW.
- Then move the NW client file index (/nsr/index/server_name)
- Restart NW - the client will automatically be generated.
- Try a new manual backup
- Try a new automatic backup
- If all runs fine, recover the client file index using "nsrck -L7 server_name"

 

[logicacmg000]

I tried running a manual backup and it fails. This is the error I get.

Backup: RPC error: Unknown host.
Backup: Cannot open Backup session with "Servername"

Should I try what you suggested about shutting down NW and moving the index file for the server?

 

[605]

Yes, please. You could go back at any time.

However, "unknown host" points to a name-to-ip resolution conflict.

 

[logicacmg000]

So I stop NW and moved the client's index file & then tried to do a manual backup but I get the same error as I was getting before (as stated above).

I am thinking should I uninstall the NW client and then reinstall it? Do you think that is worth giving a shot?

Thanks.

 

[logicacmg000]

When I do a automatic backup after restarting NW with a newly generated client index file, I get the following error.

--- Unsuccessful Save Sets ---

* "Servername":All 1 retry attempted
* "Servername":All nsrexec: authtype nsrexec
* "Servername":All savefs: unknown host name: "Servername"

 

[605]

Before the manual backup does not work, it is not worth to try an automatic one.

If you can afford to delete the old data, you may start all over - you don't even have to reinstall NW.

Simply
- stop all daemons
- delete these directories:
/nsr/index
/nsr/logs
/nsr/mm
/nsr/res
/nsr/tmp
- and restart the daemons

It should come up with a fresh installation.

 

[logicacmg000]

I think I would be much more comfortable to just uninstall this client & re-install it rather than do what you have suggested. Just my comfort level using NW.

Thanks for all your help.

 

[logicacmg000]

I un-installed NW & then re-installed it on the client. After that I started the manual backup and that failed. Later I started the automatic backup and that failed as well.

So should I try this now?

- stop all daemons
- delete these directories:
/nsr/index
/nsr/logs
/nsr/mm
/nsr/res
/nsr/tmp
- and restart the daemons

 

[605]

When you uninstalled the software, NW leaves the /nsr tree behind. As long as you did not delete it, you did not change anything.

 

[logicacmg000]

Okay here is my situation.

Since this was a production server I had to schedule a down-time. What I did was un-installed it again, and then rebooted my win2003 client and then re-installed NW 6.1.4.
After this, I was able to complete a manual backup, however, when I tried to schedule the automatic backup from the NW I still get the same error. Somewhere in NW its got the old host that use to be on the same subnet as the old IP of this client.

I am really very hesitant to try what you have suggested because if anything goes than I wouldn't know how to fix it.

 

[605]

NW does not store database information anywhere else than in
/nsr/index
/nsr/mm
/nsr/res
As long as you did not forget to delete these docs, you should not have any problem.

And you can always copy these files to a different location. In case of problems, just copy them back, restart the daemons and you are back again.

However, when manual backups are possible, the client name
could be resolved (otherwise the client resource would not have been created) and the index could be filled.

As an automatic backup is nothing else but a manual backup remotely executed from the NW server the problem must be within the start phase. The only important file is ..\nsr\res\servers at the CLIENT. It defines, which NW servers are authorized to backup this client. Please check and/or edit the file (you may even delete it). Then do not forget to restart the nsrexecd service at the client to make a change become effective.

If this works, then you must have done changes to the NW server's IP address as well. Or you must have edited the servers file.

What worries me is the error message which should be different in this case. But let's put this aside for the moment.