Hello Torsten,

thank you for your hints. I have found the user task that caused the 
shutdown, but executing the same series of select statements did not 
cause a shutdown. So I think a recovery will not help. 

Toni


Strahl, Torsten wrote:

>Hi Toni,
>this means that you have at least one record with
>a corrupted record length. This looks like a hardware 
>problem. Each select on the same record/data page will 
>fail, therefore you have to execute a data base recovery.
>
>Unfortunately the error output into the knldiag is very
>poor. I will change it in the next release, so that the
>corrupted page will be dumped and the root of the 
>corresponding b*tree will be written into the knldiag.
>
>Regards,
>Torsten
>
>SAP DB, SAP Labs Berlin
>
>-----Original Message-----
>From: Toni Epple [mailto:[EMAIL PROTECTED]]
>Sent: Montag, 14. Oktober 2002 12:20
>To: [EMAIL PROTECTED]
>Subject: 52050 SHUTDOWN select/illegal_entrylen
>
>
>Hi to all,
>
>today my database (7.3.0.20, Linux) did an emergency shutdown. It seems 
>that a select statement has been the cause (there is no other user 
>activity).
>knldiag.err says:
>8<------------------------------------>8
>2002-10-14 08:21:06   878 ERR 52050 SHUTDOWN select/illegal_entrylen
>2002-10-14 08:21:07   878 ERR 52050 SHUTDOWN *** EMERGENCY ***: 9070
>2002-10-14 08:21:31   873 ERR 11223 vfwrite   EOF after 4096 bytes
>8<------------------------------------>8
>
>The database is up and running now after a restart. Does anybody have an 
>idea what this means, and how I can prevent this from happening?
>
>Thanks
>
>Toni Epple
>
>p.s.:
>knldiag says:
>8<------------------------------------>8
>+++++++++++++++++++++++++++++++++++++++ Kernel Exit 
>++++++++++++++++++++++++++++
>2002-10-14 08:21:31   862     12845 DBSTATE  Kernel exited without core 
>and exit status 0x200
>2002-10-14 08:21:31   862     12849 DBSTATE  Kernel exited exit code 2
>2002-10-14 08:21:31   862     12808 CRASH    Creating tracewriter 
>emergency dump
>2002-10-14 08:21:31   862     11987 dump_rte rtedump written to file 
>'rtedump'
>2002-10-14 08:21:31   862     12696 DBSTATE  Change DbState to 
>'KILL_ALL'(28)
>2002-10-14 08:21:31   862     12696 DBSTATE  Change DbState to 'OFFLINE 
>'(28)
>--------------------------------------- current write position 
>-----------------
>8<------------------------------------>8
>...
>8<------------------------------------>8
>2002-10-14 08:21:06   878     11007 COMMUNIC wait for connection T30
>2002-10-14 08:21:06   878 ERR 52050 SHUTDOWN select/illegal_entrylen
>2002-10-14 08:21:07   878     11000 vdetach  '/sapdb2/EL_MUHOS/DISKS01' 
>devno 1 T29
>2002-10-14 08:21:07   871     12822 TASKING  Thread 883 joining
>2002-10-14 08:21:07   883     11566 stop     DEVi stopped
>2002-10-14 08:21:07   871     12822 TASKING  Thread 884 joining
>2002-10-14 08:21:07   884     11566 stop     DEVi stopped
>2002-10-14 08:21:07   878     11000 vdetach  
>'/sapdb2/EL_MUHOS/DISKD0001' devno 2 T29
>2002-10-14 08:21:07   871     12822 TASKING  Thread 885 joining
>2002-10-14 08:21:07   885     11566 stop     DEVi stopped
>2002-10-14 08:21:07   871     12822 TASKING  Thread 886 joining
>2002-10-14 08:21:07   886     11566 stop     DEVi stopped
>2002-10-14 08:21:07   878     11000 vdetach  '/sapdb2/EL_MUHOS/DISKL001' 
>devno 3 T29
>2002-10-14 08:21:07   871     12822 TASKING  Thread 887 joining
>2002-10-14 08:21:07   887     11566 stop     DEVi stopped
>2002-10-14 08:21:07   871     12822 TASKING  Thread 888 joining
>2002-10-14 08:21:07   888     11566 stop     DEVi stopped
>2002-10-14 08:21:07   878 ERR 52050 SHUTDOWN *** EMERGENCY ***: 9070
>2002-10-14 08:21:07   878     12621 DBSTATE  Shutdown kill requested
>2002-10-14 08:21:07   878     12696 DBSTATE  Change DbState to 
>'SHUTDOWNKILL'(26)
>2002-10-14 08:21:07   878     12696 DBSTATE  Change DbState to 'KILL'(28)
>2002-10-14 08:21:07   878     12697 DBSTATE  Resuming tracewriter
>2002-10-14 08:21:07   878     12825 TASKING  state 28 before shutkill(1)
>2002-10-14 08:21:07   878 WNG 11824 COMMUNIC Releasing  T29 database 
>shutdown
>2002-10-14 08:21:07   878     12768          UKT6 stopped
>2002-10-14 08:21:07   873     12825 TASKING  state 28 before shutkill(1)
>2002-10-14 08:21:07   866     12696 DBSTATE  Change DbState to 
>'TRACE_WRITER_WAIT'(28)
>2002-10-14 08:21:07   873     12825 TASKING  state 28 before shutkill(1)
>2002-10-14 08:21:13   877     12825 TASKING  state 28 before shutkill(1)
>2002-10-14 08:21:13   877     12768          UKT5 stopped
>2002-10-14 08:21:31   873 ERR 11223 vfwrite   EOF after 4096 bytes
>2002-10-14 08:21:31   873     12619 TASKING  Releasing tracewriter
>2002-10-14 08:21:31   863     12624 EXCEPT   Kernel aborts
>8<------------------------------------>8
>
>the kerneltrace says:
>************************************************************************
>***     KERNPROT  7.3.0                     2002-10-14  11:52:32     ***
>************************************************************************
> 
>Input File: /var/opt/sapdb/indep_data/wrk/EL_MUHOS/knltrace.dat
> 
>---------- from entry 1.6 to 1.8197 (from page 2 to 2) -----------------
> 
>===== T29 ===== id4290322/1 ================================1.6  page 2
> 
>**************************   2002-10-14 08:21:06
>**  EMERGENCY SHUTDOWN  **   select nil
>**************************   illegal_entrylen
> 
> *** opmsg: SHUTDOWN select/illegal_entrylen
> *** opmsg: SHUTDOWN *** EMERGENCY ***: 9070
>
>
>_______________________________________________
>sapdb.general mailing list
>[EMAIL PROTECTED]
>http://listserv.sap.com/mailman/listinfo/sapdb.general
>
>
>  
>


-- 

Anton Epple 

Genomatix Software GmbH
Landsberger Strasse 6
D-80339 Muenchen

http://www.genomatix.de


_______________________________________________
sapdb.general mailing list
[EMAIL PROTECTED]
http://listserv.sap.com/mailman/listinfo/sapdb.general

Reply via email to