topic Re: How to handle the INDEX db for Legato;handling /nsr file system in Operating System - HP-UX

How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Thu, 04 Oct 2007 08:04:32 GMT

HPUX-11.11/A500-5X/
Legato Networker /5.5.Build.38

how to trim the INDEX db file systesm (/nsr) . we are not in a position to increase the file system due to space constraint.

/nsr/index/hostname/db

One day i noticed the sr.0 file was taking the 95% of space. And the backup failed as the usage became 100% full. I moved that sr.0 file to temprary location to avoid the 100% usage.it again failed as the file i moved is the index db file.

Any suggestions..

Re: How to handle the INDEX db for Legato;handling /nsr file system

Robert-Jan Goossens_1 — Thu, 04 Oct 2007 08:48:09 GMT

Hi,

Check the man page of nsrck (-F) of your version.

# nsrck -F client
nsrck -F Compress Indexes

One additinional point I remember, you should have enough space in your /nsr filesystem as your largest client index database. Nsrck will rebuild the index.

Regards,
Robert-Jan

Re: How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Thu, 04 Oct 2007 12:26:33 GMT

an exerpt from from nsradmin(given below) and if i got it correctly my index databse does not HOLD information older to 15 days.

browse policy: Indexretention;
retention policy: Indexretention;

type: NSR policy;
name: Indexretention;
period: Days;
number of periods: 15;

In this case is this "compress index" operation going to help me?

Re: How to handle the INDEX db for Legato;handling /nsr file system

Robert-Jan Goossens_1 — Thu, 04 Oct 2007 15:38:55 GMT

the retention period is just the watermark for the software to allow tapes to be recycled.

If you are NOT recycling tapes (overwriting the old tapes) the media/index database will get "polluted".

Query your networker index/media databases with nsrinfo or mminfo, try to find out if your databases need to be cleaned.

---
nsrck

Use the -F option to force a check on the listed client names. If no names are given, forced checks are performed for all client indexes. This option forces all phases of index checking. For backward compatibility, the -F option implies index compression, and may be used to force the index to be compressed. This option is typically only necessary when the browse policy is reduced (for example, if the browse policy is changed from 1 year to 6 months). See "NSR policy " for information on the policy resource.
---

Besides above cleanup of the databases, there are not a lot of choices, maybe you could split /nsr over multiple filesystems.

Re: How to handle the INDEX db for Legato;handling /nsr file system

John Payne_2 — Thu, 04 Oct 2007 15:45:29 GMT

Any way you can change the retention policy on some of your backups? We had problems with the backup jockeys trying to keep large time periods of backups on our machines, even for filesystems that never changed. We would have 5, 10, 15 or more copies of full backups on filesystems that really didn't need more than 1 or possibly 2.

We cleaned things up, and saved tons of space. (Just by getting rid of the indexes of older stuff we really didn't need to keep.

hope it helps

john

Re: How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Thu, 04 Oct 2007 19:02:48 GMT

hostname:root [/opt/networker/bin] recover
recover: Using hostname as server for hostname
recover: Cannot start session with server hostname: bad database header

here is the error in /nsr/logs/daemon.log when performed "nsrck -F hostname"

10/04/07 13:38:17 nsrck: checking index for hostname
10/04/07 13:38:17 nsrd: index notice: cross-checking index for hostname
10/04/07 13:38:17 nsrck: cross-checking index for hostname
10/04/07 13:39:20 nsrck: completed checking 1 client(s)
10/04/07 13:39:20 nsrd: index notice: completed checking 1 client(s)
10/04/07 13:52:48 nsrd: index notice: checking index for hostname
10/04/07 13:55:53 nsrd: index notice: checking index for hostname
10/04/07 14:01:22 nsrd: index notice: more space needed to compress hostname index, 2098 MB required
10/04/07 15:33:09 nsrindexd: Rebuilding btrees...
10/04/07 15:39:45 nsrindexd: error on open of recfile for index /nsr/index/hostname/db (I/O error)
10/04/07 15:39:45 nsrindexd: session start error for hostname: -5

other error coming in

10/04/07 17:33:09 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_i0
10/04/07 17:33:09 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t0
10/04/07 17:33:09 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t1
10/04/07 17:33:09 nsrindexd: session start error for adeda42n.ami.alcoa.com: -525
10/04/07 19:33:09 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_i0
10/04/07 19:33:09 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t0
10/04/07 19:33:09 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t1
10/04/07 19:33:09 nsrindexd: session start error for adeda42n.ami.alcoa.com: -525
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_i0
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t0
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t1
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t2
10/04/07 19:53:02 nsrindexd: session start error for adeda42n.ami.alcoa.com: -525
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_i0
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t0
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t1
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t2
10/04/07 19:53:02 nsrindexd: session start error for adeda42n.ami.alcoa.com: -525

here is what i have in db

hostname:root [/nsr/index/hostname/db] ls -rlt
total 4714976
-rw------- 1 root root 0 Sep 26 00:53 crossck
-rw------- 1 root root 246664192 Oct 3 00:51 sr.1
-rw------- 1 root sys 2147475456 Oct 4 15:33 sr.0
-rw------- 1 root sys 8192 Oct 4 15:39 VolHdr
-rw------- 1 root sys 0 Oct 4 15:39 sr_t2
-rw------- 1 root sys 65536 Oct 4 15:39 sr
-rw------- 1 root sys 19837952 Oct 4 15:39 sr_t2.0

Re: How to handle the INDEX db for Legato;handling /nsr file system

Robert-Jan Goossens_1 — Fri, 05 Oct 2007 04:09:56 GMT

first error.

10/04/07 14:01:22 nsrd: index notice: more space needed to compress hostname index, 2098 MB required

You need more space in this filesystem to compress this client.

10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_i0
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t0
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t1
10/04/07 19:53:02 nsrindexd: Warning: /nsr/index/adeda42n.ami.alcoa.com/db missing file sr_t2

These error messages could be the result from the low space error above, otherwise I suggest to recreate your databases and use the scanner command to recover these databases from your latest tape.

Re: How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Fri, 05 Oct 2007 14:38:57 GMT

otherwise I suggest to recreate your databases -- How?

and use the scanner command to recover these databases from your latest tape.

you mean use the scanner and later use the mmrecv? And use the newly created db there after.(mmrecov will over write the existing /nsr/index/hostname/db)

I dont have any files currently under /opt/local/unix/logs/backup/bootstrap.

Re: How to handle the INDEX db for Legato;handling /nsr file system

Robert-Jan Goossens_1 — Sat, 06 Oct 2007 05:19:32 GMT

Use nwrecover to restore the clients /nsr/index/clients/client_name index databases.

Re: How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Mon, 08 Oct 2007 15:01:27 GMT

i could recover using the scanner.Now the index db is OK and backups are wokring fine.

But i dont see any files getting generated in the bootstrap folder? it has been like that earlier too.How to enable that?

Re: How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Mon, 08 Oct 2007 15:01:40 GMT

i could recover using the scanner.Now the index db is OK and backups are wokring fine.

But i dont see any files getting generated in the bootstrap folder.It has been like that earlier too.How to enable that?

Re: How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Fri, 12 Oct 2007 14:56:12 GMT

forgot to mention mmrecov was done after the "scanner" operation.

Re: How to handle the INDEX db for Legato;handling /nsr file system

Robert-Jan Goossens_1 — Fri, 12 Oct 2007 15:20:41 GMT

Hi,

This isn't your version, but it should be the same. Have a look at the pages 128/131.

http://software.emc.com/collateral/storage_software/networker_family/lssv/lssvag_61.pdf

Regards,
Robert-Jan

Re: How to handle the INDEX db for Legato;handling /nsr file system

skt_skt — Sun, 18 Nov 2007 19:25:02 GMT

closig with the current informations. Thanks to all.