We have an ontape backups from a IDS 7.31.UD8X7 with AIX 4.3.3.

For some reason we need to restore this on IDS 7.31.UD8X7 with AIX 5.3.

For a instance with small databases (< 10 GB) it worked fine.
Then we tried for a instance with bigger databases (> 50 GB).

Here the instance went down:
13:57:21 Physical Restore of rootdbs, physdbs, logdbs, svvs001,
svvs002, svvs003, svvs004, svvs005, svvs006, svvs007, svvs008
Completed.
13:57:21 Checkpoint Completed: duration was 0 seconds.
13:57:21 Checkpoint loguniq 33654, logpos 0x2d018

13:58:06 No logical log restore will be performed.
13:58:06 Clearing the physical and logical logs has started
14:39:00 Cleared 7999 MB of the physical and logical logs in 2454
seconds
14:39:00 Physical Recovery Started.
14:39:00 Physical Recovery Complete: 0 Pages Restored.
14:39:00 Logical Recovery Started.
14:39:03 Logical Recovery Complete.
0 Committed, 0 Rolled Back, 0 Open, 0 Bad Locks

14:39:04 Bringing system to Quiescent Mode with no Logical Restore.

14:39:05 Quiescent Mode
14:39:05 Assert Failed: Dynamic Server must abort
14:39:05 IBM Informix Dynamic Server Version 7.31.UD8X7
14:39:05 Who: Session(1, informix@ffzd0xa6, 0, 1342709784)
Thread(7, main_loop(), 50058018, 3)
File: rslog.c Line: 3271
14:39:05 Results: Dynamic Server must abort
14:39:05 Action: Reinitialize shared memory
14:40:53 See Also: /opt/informix/IDS.7.31.UD8/tmp/af.3ef72f9,
shmem.3ef72f9.0
14:40:53 Error writing
'/opt/informix/IDS.7.31.UD8/tmp/shmem.3ef72f9.0' errno = 28
14:40:53 rslog.c, line 3271, thread 7, proc id 950330, Dynamic Server
must abort.
14:40:53 PANIC: Attempting to bring system down

I heard it's officially not supported. But I can't see the reason for
this ?
Isn't AIX 5.3 binary compatible to AIX 4.3.3 ???

Thanks,
Jürgen

Jürgen Buck wrote:

Maybe it's not a version/OS problem:

Errno 28 on AIX is 'No space left on device!

Check that the chunk links point to partitions large enough to hold the
original chunk data.

Art S. Kagel


> We have an ontape backups from a IDS 7.31.UD8X7 with AIX 4.3.3.
>
> For some reason we need to restore this on IDS 7.31.UD8X7 with AIX 5.3.
>
> For a instance with small databases (< 10 GB) it worked fine.
> Then we tried for a instance with bigger databases (> 50 GB).
>
> Here the instance went down:
> 13:57:21 Physical Restore of rootdbs, physdbs, logdbs, svvs001,
> svvs002, svvs003, svvs004, svvs005, svvs006, svvs007, svvs008
> Completed.
> 13:57:21 Checkpoint Completed: duration was 0 seconds.
> 13:57:21 Checkpoint loguniq 33654, logpos 0x2d018
>
> 13:58:06 No logical log restore will be performed.
> 13:58:06 Clearing the physical and logical logs has started
> 14:39:00 Cleared 7999 MB of the physical and logical logs in 2454
> seconds
> 14:39:00 Physical Recovery Started.
> 14:39:00 Physical Recovery Complete: 0 Pages Restored.
> 14:39:00 Logical Recovery Started.
> 14:39:03 Logical Recovery Complete.
> 0 Committed, 0 Rolled Back, 0 Open, 0 Bad Locks
>
> 14:39:04 Bringing system to Quiescent Mode with no Logical Restore.
>
> 14:39:05 Quiescent Mode
> 14:39:05 Assert Failed: Dynamic Server must abort
> 14:39:05 IBM Informix Dynamic Server Version 7.31.UD8X7
> 14:39:05 Who: Session(1, informix@ffzd0xa6, 0, 1342709784)
> Thread(7, main_loop(), 50058018, 3)
> File: rslog.c Line: 3271
> 14:39:05 Results: Dynamic Server must abort
> 14:39:05 Action: Reinitialize shared memory
> 14:40:53 See Also: /opt/informix/IDS.7.31.UD8/tmp/af.3ef72f9,
> shmem.3ef72f9.0
> 14:40:53 Error writing
> '/opt/informix/IDS.7.31.UD8/tmp/shmem.3ef72f9.0' errno = 28
> 14:40:53 rslog.c, line 3271, thread 7, proc id 950330, Dynamic Server
> must abort.
> 14:40:53 PANIC: Attempting to bring system down
>
> I heard it's officially not supported. But I can't see the reason for
> this ?
> Isn't AIX 5.3 binary compatible to AIX 4.3.3 ???
>
> Thanks,
> Jürgen
>

The error number -28 is from the shared memory dumps.
After we repeated the restore a few times the file system for the dumps
was full.
But the chunks are located somewhere else. Sorry for sending this
confusing part of the message log.

After this we made a new backup/restore and then it worked.
But this makes it even more confusing for me, because it seems to be a
sporadic problem.

If it is not officially supported, it makes probably no sense to get of
the bottom of this problem.