Group tries to save in a different pool

[Iago77]

We have scheduled a full backup from Monday to Friday to backup Exchange Cluster.

GroupName: WindowsExchange
Clients: sexch001, sexch002, sexch006, smail1

Pool: ismLTO3

sexch001 and sexch002 are nodes, sexch006 is the quorum and smail1 is the cluster virtual manchine.

From time to time, I receive this error in daemon.log:

10/08/07 21:54:11 nsrd: nsrjb notice: nsrjb -j L700 -O504 -l -R -M -J sexch001.oepm.local 023370
10/08/07 21:54:13 nsrd: media info: loading volume 023370 into rd=sexch001.oepm.local:\\.\Tape0
10/08/07 21:54:13 nsrd: Jukebox 'L700' failed: Illegal Request, Incompatible Medium Installed

23370 is a volume from pool Default so I don't understand why this one uses a tape from another pool (ismLTO3 pool has enough free tapes).

Besides, it always happens with one of the nodes. This is the backup log of the group:

[...]

Start time: Mon Oct 8 20:30:00 2007
End time: Tue Oct 9 00:51:00 2007

--- Unsuccessful Save Sets ---

* sexch001: VSS (returned from savefs) skipped.
The saveset returned from savefs probe does not match the one specified for the client.

[...]

Due to an obscure reason, Legato tries to save the saveset within another pool which is not specified to this group.

Could you give me some advice?

Many thanks again,

Iago

 

[605]

It looks like there are 2 (potential) problems:

A media database problem could cause the 'incompatible medium installed' problem. In this case you might want to delete/relabel the tape.

A configuration problem for your groups/pools/storage node lists. This is more difficult to examine as we do not have this info.

 

[Iago77]

23370 tape was ismLTO3, so that's the reason why WindowsExchange group couldn't write to this one.

Anyway, what's the cause related? WindowsExchange takes another volume from Default pool

 

[Iago77]

I'll try to explain to you our arquitecture:

Group: WindowsExchange
Pool: ismLTO2
Devices though StorageNode: /dev/rmt/1cbn, 2cbn and 3cbn
StartTime: 20:30

Group copies successfully all the clients but one: sexch001.

The daemon.log shows:

10/08/07 20:30:58 nsrd: media waiting event: Waiting for 1 writable volumes to backup pool 'ismLTO2' tape(s) on sexch002.oepm.local
10/08/07 20:30:59 nsrd: media info: suggest relabeling 019104 on sexch001.oepm.local for writing to pool 'Default'

This sounds weird, sexch001 is asking for a tape from Default pool. According to my experience, Default pool is used if: a) pool is not specified; b) specified pool has no volumes left.

a) Pool is specified in the pool definition:

# nsradmin

nsradmin> print type:NSR pool; name:ismLTO2
type: NSR pool;
name: ismLTO2;
comment: ;
enabled: Yes;
pool type: Backup;
label template: codigo;
groups: Temporal_ismLTO2, WindowsDatos, WindowsExchange,
WindowsExchangeMB, WindowsFicheros, WindowsSO;
clients: ;
save sets: ;
levels: ;
devices: /dev/rmt/1cbn, /dev/rmt/2cbn, /dev/rmt/3cbn,
"rd=atenas:\\\\.\\Tape1",
"rd=atenas:\\\\.\\Tape2",
"rd=sefic001.oepm.local:\\\\.\\Tape0",
"rd=sefic001.oepm.local:\\\\.\\Tape1",
"rd=sefic001.oepm.local:\\\\.\\Tape2",
"rd=sefic002.oepm.local:\\\\.\\Tape0",
"rd=sefic002.oepm.local:\\\\.\\Tape1",
"rd=sefic002.oepm.local:\\\\.\\Tape2",
"rd=sexch001.oepm.local:\\\\.\\Tape0",
"rd=sexch001.oepm.local:\\\\.\\Tape1",
"rd=sexch001.oepm.local:\\\\.\\Tape2",
"rd=sexch002.oepm.local:\\\\.\\Tape0",
"rd=sexch002.oepm.local:\\\\.\\Tape1",
"rd=sexch002.oepm.local:\\\\.\\Tape2";
store index entries: Yes;
auto media verify: No;
Recycle to other pools: No;
Recycle from other pools: No;
volume type preference: ;
mount class: default;

b) There are enough free volumes:

# mminfo -m -q 'location=L700,pool=ismLTO2,volrecycle' | wc -l
72

I have realised there's a message which may be related to this issue:

10/08/07 20:30:50 nsrd: Jukebox 'L700' failed: All of the devices are in use by nsrmmd

Until any of the devices is free, relabelling operation is delayed. Once there's a spare device:

10/08/07 22:39:25 nsrd: media waiting event: Waiting for 1 writable volumes to backup pool 'Default' tape(s) on sexch001.oepm.local
10/08/07 22:39:25 nsrd: nsrjb notice: nsrjb -j L700 -O539 -l -R -M -J sexch001.oepm.local 023370
10/08/07 22:39:26 nsrd: media info: loading volume 023370 into rd=sexch001.oepm.local:\\.\Tape0
10/08/07 22:39:27 nsrd: Jukebox 'L700' failed: Illegal Request, Incompatible Medium Installed

23270 is a tape belonging to Default pool. 23370 is LTO3 so Legato can't complete the label operation.

I think the point is about guessing why Legato is using a Default tape.

Many thanks for your help,

Iago

 

[605]

No real clue. The pool's setup seems to be o.k. However, it works like it is disabled.

At the moment i can only suggest that you open a case with support.