serious volume problem
2015-10-04 17:46:05
I think a have a serious problem. Well, actually I don't think so, but
I'am sure.
A volume (U00020) is defined to a storagepool (UNITEPOOL) and is
used for backup's. After a while, Space reclamation started for this
volume:
03/29/99 09:33:53 ANR1040I Space reclamation started for volume U00020,
storage pool UNITEPOOL (process number 597).
03/29/99 09:33:53 ANR1044I Removable volume U00020 is required for space
reclamation.
03/29/99 09:33:53 ANR8324I 3590 volume U00020 is expected to be mounted
(R/O).
03/29/99 09:33:53 ANR1142I Moving data for collocation cluster 1 of 1 on
volume U00020.
03/29/99 12:28:20 ANR1175W Volume U00020 contains files which could not
be
reclaimed.
03/29/99 12:28:20 ANR1410W Access mode for volume U00020 now set to
"unavailable".
03/29/99 12:28:20 ANR1041I Space reclamation ended for volume U00020.
Well, after the space reclamation ended and the volume is set to
"unavailable",
it is disappeared from the storagepool:
adsm> q vol U*
....
U00018 UNITEPOOL 3494-3590 0.0 0.0 Pending
U00019 UNITEPOOL 3494-3590 0.0 0.0 Pending
U00021 UNITEPOOL 3494-3590 14,920.0 68.1 Full
U00022 UNITEPOOL 3494-3590 0.0 0.0 Pending
....
Every morning at 09:00, a backup db started, which should use only
scratch volumes. Well, at least, this is the way it works for a long
time :-(
SURPRISE:
Private volume U00020 is used:
03/30/99 09:00:24 ANR2750I Starting scheduled command BACKUP_DB ( backup
db
devclass=3494-3590 type=full ).
03/30/99 09:00:24 ANR2017I Administrator ADMIN issued command: BACKUP DB
devclass=3494-3590 type=full
03/30/99 09:00:24 ANR0984I Process 628 for DATABASE BACKUP started in the
BACKGROUND at 09:00:24.
03/30/99 09:00:24 ANR2280I Full database backup started as process 628.
03/30/99 09:00:24 ANR2280I Full database backup started as process 628.
03/30/99 09:00:24 ANR2753I (BACKUP_DB):ANR2280I Full database backup
started
as
03/30/99 09:00:24 ANR2753I (BACKUP_DB):process 628.
03/30/99 09:00:24 ANR2756I Scheduled command BACKUP_DB started
successfully.
03/30/99 09:01:05 ANR8337I 3590 volume U00020 mounted in drive 3590-2
(/dev/rmt2).
03/30/99 09:01:07 ANR1360I Output volume U00020 opened (sequence number
1).
.....
.....
03/30/99 09:52:40 ANR4554I Backed up 2292528 of 2326863 database pages.
03/30/99 09:53:10 ANR4554I Backed up 2316736 of 2326863 database pages.
03/30/99 09:53:31 ANR1361I Output volume U00020 closed.
03/30/99 09:53:33 ANR8336I Verifying label of 3590 volume U00020 in drive
3590-2 (/dev/rmt2).
03/30/99 09:53:36 ANR4550I Full database backup (process 628) complete,
2326863 pages copied.
03/30/99 09:53:36 ANR0985I Process 628 for DATABASE BACKUP running in the
BACKGROUND completed with completion state SUCCESS at
09:53:36.
03/30/99 09:54:02 ANR8468I 3590 volume U00020 dismounted from drive
3590-2
(/dev/rmt2) in library 3494LIB.
And volume U00020 suddenly appear in the libvol as Private/DbBackup:
adsm> q libvol * U*
3494LIB U00018 Private Data
3494LIB U00019 Private Data
3494LIB U00020 Private DbBackup
3494LIB U00021 Private Data
3494LIB U00022 Private Data
So, the backup-data from volume U00020 is gone for ever.
And this is not the only volume which is destroyed this way.
Questions:
1. Why are the volumes disappear from the storagepool?
2. Why is backup db using private volumes?
Anyone any clues?
Cheers,
Henk ten Have ( ADSM Server 3.1.2.15 on AIX 4.2.1.0)
|
|
|