Re: Single row update - assert failed
Get rid of your in place alter tables;
sounds like you hit a bug with in place alter tables
i do not have the sql at hand; search for it or maybe someone else can send
you the sql to find it; eq identify which tables have i.p.a.
Using oncheck it is oncheck -pT <dbname> or per table
oncheck -pT <dbname>:<table>
search for
Home Data Page Version Summary
Version Count
0 (oldest) 0
1 0
2 (current) 3
if you find it do a dummy update or without logging
alter fragment on <table> init in <dbspace>;or reload or....
See you
Superboer.
MF <aussiewoof@aol.com> wrote in message news:<Xns9581A891D8EF0aussiewoofaolcom@140.99.99.130>...
> Hi ,
>
> I'm using Informix 7.30 on a solaris 9 box and it recently crashed after
> a single row update with the following error:
>
> 14:35:58 fatal undo-rowalter error: rowsize(72) exceeding space> available(66) in orignal page
> 14:35:58 fatal undo-pgalter error: undo_rowalter() failed
> ...
> 14:35:59 Assert Failed: Page Check Error in undopgalter:undo_rowalter> failed
> 14:35:59 Who: Session(35, root@lsbprod, 0, 0)
> Thread(136, xchg_1.3, 0, 1)
> File: rsdebug.c Line: 943
> 14:35:59 Results: Possible inconsistencies in 'lsb:"".'
> 14:35:59 Action: Run 'oncheck -cD 2097203'
> 14:35:59 sh /opt/informix/etc/evidence.sh 2 0 /tmp/af.88de0e 35 0x0 136> 0x8a340880 1 0 0 0 0
> 14:35:59 See Also: /tmp/af.88de0e, shmem.88de0e.0
> 14:35:59 Error writing '/tmp/shmem.88de0e.0' errno = 22
> 14:35:59
> ------------------ End of assertion failure 0 ----------------->
> 14:35:59
> 14:35:59 Assert Failed: Dynamic Server must abort
> 14:35:59 Who: Session(35, root@lsbprod, 0, 0)
> Thread(136, xchg_1.3, 0, 1)
> File: called by ASF mt_affail Line: 0
> 14:35:59 Results: Fatal Internal Error requires system shutdown
> 14:35:59 Action: Restart OnLine
> 14:35:59 sh /opt/informix/etc/evidence.sh 2 0 /tmp/af.88de0e 35 0x0 136> 0x8a340880 1 0 0 0 0
> 14:35:59 See Also: /tmp/af.88de0e, shmem.88de0e.1
> 14:35:59 Error writing '/tmp/shmem.88de0e.1' errno = 22
> 14:35:59
> ------------------ End of assertion failure 1 ----------------->
> Unfortunately the system could not perform a fast recovery as it would
> crash when restoring the logical logs. As the system could not enter
> Quiescent mode there was no option to run oncheck so the system had to
> be completely restored from tape. Any ideas as to why this would have
> happened (is it just a random inconsistency that required oncheck) and
> what solution other than restore the instance from backup could I have
> followed.
>
> Thanks for any ideas
> Mark
>
>
> PS - I know I need to upgrade
>
>
> This is the output from attempting to restart
>
> 17:26:53 Physical Recovery Started.
> 17:26:53 Physical Recovery Complete: 384 Pages Restored.
> 17:26:53 Logical Recovery Started.
> 17:26:54 Assert Failed: Page Check Error in undopgalter:undo_rowalter> failed
> 17:26:54 Informix Dynamic Server Version 7.30.UC2
> 17:26:54 Who: Session(30, informix@lsbprod, 0, 0)
> Thread(136, xchg_1.3, 0, 3)
> File: rsdebug.c Line: 943
> 17:26:54 Results: Possible inconsistencies in 'lsb:"".'
> 17:26:54 Action: Run 'oncheck -cD 2097203'
> 17:26:58 See Also: /tmp/af.88061d, shmem.88061d.0
> 17:26:58 Error writing '/tmp/shmem.88061d.0' errno = 22
> 17:26:58 rstrans.c, line 7773, thread 136, proc id 3774, Dynamic Server> must abort.
> 17:26:58 Assert Failed: Dynamic Server must abort
> 17:26:58 Informix Dynamic Server Version 7.30.UC2
> 17:26:58 Who: Session(30, informix@lsbprod, 0, 0)
> Thread(136, xchg_1.3, 0, 3)
> File: called by ASF mt_affail Line: 0
> 17:26:58 Results: Fatal Internal Error requires system shutdown
> 17:26:58 Action: Restart OnLine
> 17:27:03 See Also: /tmp/af.88061d, shmem.88061d.1
> 17:27:03 Error writing '/tmp/shmem.88061d.1' errno = 22
> 17:27:03 rstrans.c, line 7773, thread 136, proc id 3774, Dynamic Server> must abort.
> 17:27:03 PANIC: Attempting to bring system down
> 17:48:16 Requested shared memory segment size rounded from 2096912KB to> 2097152KB
> 17:48:20 Segment locked: addr=0xa000000, size=2147483648