I have 2 Legato V 7.6.3 servers LGT1 and LGT2. Because I have to re-install SRV1 completely with Win2008-R2, I first want to move the backup information made by LGT1 to LGT2. In that way, LGT2 will backup everything (temporary) and LGT2 must be able to do restores of backups made by LGT1.
I have created on SRV2 all the legato clients using the identical client ID as on LGT1.
I than copied the index\'clients' folders from LGT1 to LGT2 when both environments are not running.
I than see on LGT2 that the size of the index of the client is the same as on LGT1
But this is not enough to be able to restore data (from backups that have been made by LGT1) with LGT2.
What do i have to do more ?
The file index is useless if you do not have the media index info. And there is currently no other way but to scan the media. You do not really want to do that.
Let me suggest you prepare your new system and then run a NW disaster recovery procedure. Just be sure that you use the same hostname. This is the much faster way. Depending on your bootstrap, this should not take longer than1-2 hours.
If I use the same hostname..... the system ID in the Windows OS will be different anyway .... That is no problem ?
What if the new hostname is different from the old one ? Can I still use the disaster recovery procedure ?
Do you have a clear description of the disaster recovery procedure ?
mmrecov -N to recover the media and resource db ? This does not overwrite the existing db ?
nsrck is not needed anymore if i manually copy the index folder from LGT1 to LGT2 using identical client IDs ?
mmrecov is the only procedure you must run for a DR. There should be a very detailled description in the docs.
Yes - this will OVERWRITE your media db. So be careful.
The only part that will be affected by the new SystemID are your auth codes (licenses). You must request new ones within 14 days.
nsrck is important to recover the old indexes (nice to have for recoveries). However, as soon as you finished the DR process, you system can run new backups.
So, if i don't want to overwrite the mm database, i have to scan tapes ? And first print out a list of tapenames and the savesets on it.... so i can find which tapes to scan
Yes, in this case, scanning is the only solution.
I understood your original message in the way that you wanted to move the WHOLE environment to a new server.
If I save the whole legato environment from Legato SRVA to a new virgin Legato SRVB, i would expect that such an operation should be possible because the index and mediadatabase are both empty on the new one.
This would lead me to the scenario:
1. create the clients of SRVA on SRVB using the identical ID.
2. stop Legato on both servers
3. copy the index and media database from SRVA to SRVB (maybe the res folder as well ??? )
4. start legato on SRVB
I think in this way, the legato environment can be migrated. Is this right?
What about the fact that the Legato ID of the servers is different ?
PS: If this does not work, i have to use the scanner command; i fear.
i am just trying to convince you to use the standard NW DR procedure which has been used for more than 20 years now. Basically, it will do the same what you want to achieve by copying the files. However, it is the supported way. The only things you need to consider are ...
- same OS family
- same hostname
- a compatible drive to recover the bootstrap from
The result is a copy of the same NW configuration and databases as before - only your auth-codes will be affected.
And the procedure is very properly documented.
Other things you need to do later are:
- adjusting the backup hardware in case something has changed here
- recovering the indexes (or just copy them from the source). As new backups will create new index directories, there will be no conflicts
- scan your AFTD media (if you used 'mmrecov -N')
- most likely the GSS certificates need to be renewed. Just delete the existing ones from the NW server
Once again, i encourage you to familiarize with the routine. So please have a look at the docs AND test the procedure - you have nothing to loose. Good luck.
Would suggest you to go with disaster recovery procedure instead of doing copy paste work, copy paste and creating client instances manually will land you up in unexpected trouble.
Follow the below procedure which is supported by EMC as said by Socca, will give u a sequence which u can follow:
1. Take bootstrap and index Backup of LGT1 on a tape which will be used for restoring on LGT2(keep bootstrap ssid details handy use mminfo -B to get bootstrap details on Backup server).
2. Install fresh Networker on LGT2 and do mmrecov and give the bootstrap ssid which needs to be restored.
3. Once the bootstrap is recovered, check for the clients and mm database on LTG2, it should be exactly same with LGT1. (Hence no need to create resources manually).
4. As this is a new hardware , you will have to rebuild the indexes of all the clients. You can use below given
command to resotre indexes from tape :
nsrck -L7 client_name.
5. Change the Backup seerver name is <installed dir>\Legato\nsr\res\servers file of all your clients so that they should point to LGT2.
6. Try doing a Backup and restore from old data once every thing is in place.
Get back to me if u get stuck anywhere.
Hope this helps you up!!!!!!!!!! Cheers
Suraj S. Pujari