Unsolved

This post is more than 5 years old

1 Rookie

 • 

47 Posts

2562

February 11th, 2009 08:00

Backup Server with two diff. client IDs

Hello to all,

This is my first post in this forum and I hope somebody from you NetWorker Gurus can help us.

We are facing with following problem

Facts:

Old backup server (NetWorker 7.3)
cy-devpl01.domain

23.11. 2007 Update to Networker 7.4

New backup server (NetWorker 7.4.Build.187 Power Edition)
cy-bkp01.domain.com (new hardware  Sun-Fire-V890, Sun Solaris 10.0)



User@cy-bkp01: >mminfo -av -r client,clientid -q client=cy-bkp01 |sort -u
client client ID
cy-bkp01 446e303c--xxxxxxxx-xxxxxxxx-xxxxxxx-xxxxxxxxx-xxxxxxxx

User@cy-bkp01: >mminfo -av -r client,clientid -q client=cy-bkp01.domain.com |sort -u
client client ID
cy-bkp01.domain.com c5373bbb-xxxxxxxx-xxxxxxxx-xxxxxxx-xxxxxxxxx-xxxxxxxx

User@cy-bkp01:/export/home/User>

User@cy-bkp01: > recover
39584: recover: No files have been backed up for this client. No index was found on the server for the path User@cy-bkp01: > on this client.
User@cy-bkp01: >


User@cy-bkp01: >mminfo -avot -c cy-bkp01

008665 cy-bkp01.domain.com 11/23/07 07:54:31 AM 27 MB 1044809527 cE full index: cxxxx
008665 cy-bkp01.domain.com 11/23/07 07:55:07 AM 37 MB 1028032347 cE full index:dxxxx
008665 cy-bkp01.domain.com 11/23/07 07:58:10 AM 15 MB 1011255314 cE full index:fxxxx
008665 cy-bkp01.domain.com 11/23/07 07:58:35 AM 26 MB 994478123 cE full index:xxxxx
007455 cy-bkp01.domain.com 11/23/07 08:00:43 AM 3333 KB 977701035 cE 9 index:qqqqqqqqqq
008665 cy-bkp01.domain.com 11/23/07 08:01:59 AM 110 MB 960923895 cE full index:eeeeeeeeeeee
008576 cy-bkp01.domain.com 11/23/07 08:14:05 AM 14 MB 608603433 cE full index:rrrrrrrrrrrrrrrrr
008576 cy-bkp01.domain.com 11/23/07 08:14:32 AM 17 MB 541494569 cE full index:fffffffffffffffffff
008576 cy-bkp01.domain.com 11/23/07 08:16:04 AM 8825 KB 256281897 cE full index:ffffffffffffffasasa
008576 cy-bkp01.domain.com 11/23/07 08:14:32 AM 17 MB 541494569 cE full index:reeeeeeeeeerrrrrrrrr
008576 cy-bkp01.domain.com 11/23/07 08:16:04 AM 8825 KB 256281897 cE full index:dsdsdsdsds
008576 cy-bkp01.domain.com 11/23/07 08:16:40 AM 33 MB 122064169 cE full index:sdddddddddd
010409 cy-bkp01 11/23/07 03:33:31 PM 2825 MB 4282839755 cE full index:2333333333333333
010409 cy-bkp01 11/23/07 03:34:51 PM 420 MB 4266062619 cE full bootstrap
010409 cy-bkp01 11/23/07 03:36:17 PM 2825 MB 4249285503 cE full index:cxxxxxxxxxxx
010409 cy-bkp01 11/23/07 03:37:08 PM 420 MB 4232508325 cE full bootstrap
010409 cy-bkp01 11/23/07 03:39:43 PM 2825 MB 4215731285 cE full index:xxxxxxxxxxxxxxxx
010409 cy-bkp01 11/23/07 03:40:32 PM 420 MB 4198954096 cE full bootstrap
010404 cy-bkp01 11/23/07 03:45:50 PM 4199 KB 4081513903 cE full /t3
010404 cy-bkp01 11/23/07 03:45:51 PM 11 GB 4064736687 cE full /local2
010404 cy-bkp01 11/23/07 03:45:52 PM 29 KB 4047959471 cE full /system/contract
010404 cy-bkp01 11/23/07 03:45:53 PM 5702 KB 4031182255 cE full /system/object
010404 cy-bkp01 11/23/07 03:45:55 PM 204 KB 4014405043 cE full /devices
010404 cy-bkp01 11/23/07 03:45:56 PM 23 GB 3997627827 cE full /local
010404 cy-bkp01 11/23/07 03:46:00 PM 1457 MB 3980850616 cE full /var
010404 cy-bkp01 11/23/07 03:48:45 PM 4671 MB 3964073565 cE full /


In short words:
Our Backup Server is having two different client IDs and currently we can not do any recovery of the backup server in case of disaster or wrong settings¿ Also the nsrck ¿L7 command does not work for data that are backed up before the 23 Nov. 2007.
Next updated to 7.4 SP 3 should be performed but first we have to solve this problem before we can make update to SP3

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 11th, 2009 13:00

No update will make this go away. Mistake has been done at early stage. I would suggest to scan older data under name of bogus client.

2 Intern

 • 

2K Posts

February 11th, 2009 20:00

Did you check with support. They would have some process for this. They had helped us with mminfo etc when we had two client ID's for the same client.

443 Posts

February 11th, 2009 21:00

It looks like you have 2 clients for the same machine. One with the short name another with the FQDN. While restoring, try using recover -c cy-bkp01.domain.com You should be able to restore the backups using this.

As far as making sure that you have only one client, you might need to delete the client resource where no index entries are getting stored. If you look at the index size you should get an idea of which one you need to delete. Once you delete the client resource, say the one with the short name, add the short name as an alias for the client resource with the FQDN.

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 12th, 2009 00:00

I think the problem here is more complex as we are talking about backup server which has two hostids - right? If not, if that would be a client, then it would be easy to fix it.

443 Posts

February 12th, 2009 01:00

You are right. Does look like the backup server. If it is the backup server, contacting support would be the best bet.

1 Rookie

 • 

47 Posts

February 12th, 2009 01:00

Unfortunately you have right Hrvoje, our problem is the Backup Server with two host IDs. The mistake was happened, obviously during the NetWorker update from 7.3 to 7.4 and now we would like to get it away before we go to 7.4.3 or 7.5.
What did you mean exactly as you said/ suggest ¿to scan older data under name of bogus client?¿

443 Posts

February 12th, 2009 02:00

Did you try to do a recover -c cy-bkp01.domain.com I had a similar problem once and I was able to recover data using the old backup server / client resource name. In terms of merging the index entries for the two different client ids, it's a little more complicated. I got the exact steps from support that time around and it was tricky even with the instructions. Would suggest a backup of your /nsr directory with networker services down before you attempt any merging.

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 12th, 2009 03:00

It really depends how much data you have prior to upgrade which should still be recoverable. If that amount is not that big (and I would assume it is not by guessing that would be just monthend backup) then I would create new clientname using old hostid first (you must create bogus entry in hosts table). Then you scan old data. -c as Hari said might do the trick and ssid restore might also work (if data is recoverable). It would be interesting to see what caused this; recently I have seen same thing upon update from 7.4SP2 to 7.5. Few tests showed that this does not happen when nsrauth is disabled (no idea why).

50 Posts

February 12th, 2009 05:00

Update to my earlier post..does nsradmin show server having two id's?

50 Posts

February 12th, 2009 05:00

I see that you have a full bootstrap backup. Maybe you can use mmrecov to recover old media DB?

Does nsradmin show the client having two clientids?

1 Rookie

 • 

47 Posts

February 12th, 2009 05:00

Regarding scan of old data,
well the company is using Legato Backup software since 2001 and amount of saved data is really big, (if we are take only the monthend backup ¿ 3 -5 LTO tapes/month it will be more than 100 LTO 1 tapes,.and in real we are having much , much more important data and actually all data that are saved should be also recoverable, even if their retention has expired, so that with scanning of all old tapes is almost impossible at the moment)

for recover ¿c option

When I try to run recover -c following message occurs:

User@cy-bkp01:/>recover
39584:recover: No files have been backed up for this client. No index was found on the server for the path / on this client.
User@cy-bkp01:/>


User@cy-bkp01:/>recover -c cy-bkp01.domain.com
39584:recover: No files have been backed up for this client. No index was found on the server for the path / on this client.
User@cy-bkp01:/>

-------------
ps.
The Service Request to our Support is going to be opened still today¿

1 Rookie

 • 

47 Posts

February 12th, 2009 06:00

how can I check it with nsradmin?

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 12th, 2009 07:00

You can't - you can see that with mminfo or from NW GUI if you check details for client (which will show current hostid being used).

I personally would not trust LTO1 to keep my data safa for that log and besides I would count only data that is recoverable. Obviously you could have two server setups; one with older state and and one with current one and switch those depending which restore date you need. Not the best solution, but after all this time you do not have many options left.

50 Posts

February 12th, 2009 09:00

nsradmin -i client_info.txt

where client_into.txt has...

-----------------
. type: nsr client

show name;comment;group;schedule;browse policy;retention policy;save set;group;schedule;backup command;application information;Aliases;Scheduled backup;networker version;ndmp;parallelism;remote access;storage nodes;client id;

print
-------------

This will show you all details of a client or server..a sample of my listing..

Current query set
name: prd-tapesrv2;
client id: \
25bec060-00000004-3ffc535a-3ffc5359-00010000-ac11645f;
-------------

etc..

6 Operator

 • 

14.4K Posts

 • 

56.2K Points

February 13th, 2009 01:00

Indeed, I stand corrected. For online nsradmin queries you can see clientid (I've been doing too many of those offline queries lately).

0 events found

No Events found!

Top