nsrmmdbd errors after nsrck -L7

[tuxedo91]

Hi,

I'm encountering nsrmmdbd errors when recovering indexes:

after nsrck -L 7 -t '02/26/2007 05:42:06',

messages are returned:
nsrmmdbd: Media DB error: Save set eligibility can not be cleared, no clone instance is within recyclable time period.

has someone has already encountered this messages?

Thanks for your replies.

 

[Rif123]

Hi,

haven't seen this before unfortunately. Your nsrck command will restore all indexes of the specified time. Is that what you actually want? Otherwise you should specify a client name after the time. What version of networker are you running?

 

[tuxedo91]

no, it's for one single client. The actual command was passed with the name of the specific client.
version networker: 7.3.2

Regards.

 

[Rif123]

This is very confusing. The error message, is that coming while you run the nsrck command? Is there any more information in the daemon.log?

Can you have a look on the status on the indexes you are trying to recover? Also the backups that the indexes will be pointing to? Are the index save sets recoverable? The data backups are they recyclable or in some other state? Are any of the save sets set to suspect? Do you get the same problem if you recover another newer index??

You can still recover the save sets if they are still available by running a save set recover. You can specify what part you want to recover and you can also redirect it. That doesn't solv the index problem but at least you should be able to recover your data.

 

[tuxedo91]

the index save sets are recoverable, and no save sets are marked suspect.
additional information: changes have been made in browse and retention policy.
the result of scanner -i gives:
nsrmmdbd : media db error ssid... retention time is invalid

 

[Rif123]

What? That's so strange. Have you changed date/time settings on the server? I don't see how that should matter, but in lack of other better suggestions...

How does it look when you run a mminfo query against this save set?
mminfo -q "ssid=1234" -r ssretent(20),savetime(20)
See anything different if you make the same query against a new save set?

What if you change the retention on the save set first then? You can do that with:

nsrmm -S 12345 -e "New retention date"
Does nsrmmdbd still give you an error while running scanner or nsrck?

How does your new policy for browse/retention look like?

 

[605]

How long did you set the retention policy? -

If you try to go beyond the 2^32 bit limit (sometime about 2037) IN NMC, you will get an error right away. And most likely nsrd will die as well.


With the old nwadmin program, there seems no problem defining such a retention policy. I also found that scanning is not a problem. However, retention date in the media db (ssretent) will show with "no date".