Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

3892

May 6th, 2009 01:00

!! URGENT: can't change browse time

Hi all,

I'm using 7.2.2 server on Windows. I need to restore some save set whose browse time is expired but retention time is ok. According to docs, I used

nsrmm -S ssid -w 5/30/2009

It gives me "nsrmm: Cannot extend browse time of a recoverable save set"

I searched through forum but no results. I'll need the indexes very urgent, please help..


Best regards,
Atatur

1.1K Posts

May 7th, 2009 03:00

It is important to distinguish here between an index and an index backup.

Your indexes are in the folder /nsr/index/clientname, when a backup runs this files in this directory are updated with contents of the new backup; when the browse period is passed the relevant files are removed.

When an index is backed up the files in /nsr/index/clientname are backed up. These remain on tape until the corresponding original backup passes out of its retention; they are then marked as recyclable.

If your backup has passed out of its browse period and the files in /nsr/index/clientname have been deleted they normally can be restored using nsrck -L7 clientname. In your case it appears that the index backup has been deleted. This means we have to make a new index rather than recover a backup of an index.

The scanner -i command will read through the tape; it will find what files are on the tape and write new files in the /nsr/index/clientname folder listing what files are on the tape which will allow us to browse through the files in the GUI or at the recover command and Networker will then be able to take our file selection and locate the relevant section of backup and read the data back in.

11 Posts

May 6th, 2009 02:00

Hi again,

Sorry for my previous hurry post. I incorrectly stated the problem; I recovered one set of indexes by using
nsrck -t "3/29/09" -L7 client1

But for the second client, I get;

nsrck: File index error: can't find index backup for client2 on server bck-srv


Best regards,
Atatur

253 Posts

May 6th, 2009 02:00

To recover index and merge them to existent ones, you should run from Backup Server:

nsrck -L7 -t "mm/dd/yyyy" client

where mm/dd/yyyy is the date by which you need to recover the index and client is the machine that has the information that you need.

253 Posts

May 6th, 2009 03:00

Could it be that there is no index for this date and client? If there is no index, you still have 2 choices:

1) scanner -i -S SSID --> to recreate the index
2) Recover by SSID

253 Posts

May 6th, 2009 04:00

mminfo -a -q "client=client2,name=index:client2" -r "volume,savetime(20)" gives you all indexes for client2. Use mminfo to know the ssid of data that you need to recover.

11 Posts

May 6th, 2009 04:00

Thanks for the answer. I'm about to recreate those indexes with your first option, but before that is there a way to see indexes for this date and client other than

mminfo -av -s server -c client

Also, from the output of above, I derived the necessary volume is ABC24L3, but from the Networker Recover GUI, the relevant save set is on ABC16L3??


Regards,
Atatur

1.1K Posts

May 6th, 2009 05:00

There seems to be a little confusion here... What has probably happened is that the saveset and index have expired, and the media with the expired index has been reused. If that is the case we are not going to recover that index because it is no longer in existence. Instead we will use the scanner command to scan the saveset and build a new index. You need to do the following:

1. Determine the saveset id of the saveset you wish to recover using mminfo, also determine the volume it is on
2. Set the saveset to notrecycable and with a retention period greater than today using nsrmm
3. After this try nsrck -L7 -t time clientname, if that fails rebuild the index:
4. Rebuild the index by loading without mounting the tape then using scanner -i -S ssid /drivepath - if it spans volumes it will reach the end and ask you to load the next volume; once this completes that saveset will be browsable

1.1K Posts

May 6th, 2009 05:00

It seems this has not passed out of its retention time, only its browse time. When it passes out of the browse time the indexes are deleted but the backups remain on media still. You can only extend the browse time if the indexes still remain.... Your index will have a retention equal to the longest retention time of backups it indexes so it will not be removed from backup media until the backup has expired, except due to external factors... nsrck -L7 clientname will recover the index for you in this case without problem...

11 Posts

May 6th, 2009 06:00

nsrck -L7 clientname failed to recover index unfortunately; giving the error above. You're correct when you're saying that browse time hasn't passed out and the savesets are in the media db..

Only thing that remains seems to apply scanner command with "-i -S ssid" options.

Thanks for your attention..

11 Posts

May 7th, 2009 00:00

Hi,

There seems 2 SSIDs relevant to my indexes and I used "scanner -i -S SSID1" to recreate index. Command succeeds but neither from Recover GUI or "nsrinfo -s server -n oracle client1" shows the supposedly recreated index.

What I'm doing wrong?

1.1K Posts

May 7th, 2009 01:00

You are not scanning in your index, you are scanning in the original saveset to build a new index.

11 Posts

May 7th, 2009 01:00

Can you be more specific, shouldn't I be seeing those indexes with nsrinfo or via GUI after scanner run? Do I need to restart networker or use nsrck -L7 or something similar??

11 Posts

May 7th, 2009 05:00

David, Sergio and all other guys, thank you very much for your valuable advices, deeply appreciated..
No Events found!

Top