Hi Holger,
The following is the dbproc code which caused this problem:
SELECT i_stock INTO :i_stock FROM dbt.item WHERE i_id=:scl_i_id;
IF (i_stock-10>scl_qty) THEN
UPDATE dbt.item set i_stock=(:i_stock-:scl_qty) where
i_id=:scl_i_id
ELSE
UPDATE dbt.item set i_stock=(:i_stock-:scl_qty+21) where
i_id=:scl_i_id;
It is interesting that if I change the code to:
SELECT i_stock INTO :i_stock FROM dbt.item WHERE i_id=:scl_i_id;
IF (i_stock-10>scl_qty) THEN
set i_stock=i_stock-scl_qty
ELSE
set i_stock=i_stock-scl_qty+21;
UPDATE dbt.item set i_stock=:i_stock where i_id=:scl_i_id
Then no 9405 error is generated. Reading from the SAPDB manual, it
seems to me that I can use this kind of expression in UPDATE, but it is
not the case, or there is a bug.
I might be asking too much, but I would like to know how you linked the
error to the statement. Is there any document about how to interpret
vtrace? It could have saved me four weeks if I knew how to read vtrace.
Thanks for your help,
Jenny
On Fri, 2002-09-27 at 08:14, Becker, Holger wrote:
> Hi Jenny,
> now we can see the problem in the vtrace.
>
> The problem occurres at these statement:
> UPDATE DBT.ITEM SET I_STOCK=(?-?) WHERE I_ID=?
>
> Unfortunately we couldn't reproduce the behaviour here.
>
> So could you please send us the code of the dbproc
> and the kernel version you use?
>
> As a workaround you could try to replace the expression
> in the update statement.
>
> Kind regards,
> Holger
> SAP Labs Berlin
>
> > -----Original Message-----
> > From: Jenny Zhang [mailto:[EMAIL PROTECTED]]
> > Sent: Freitag, 27. September 2002 01:15
> > To: Becker, Holger
> > Subject: RE: -9405 AK Datapart too short
> >
> >
> > Holger,
> >
> > I did not check the files before I posted it. Sorry for the
> > inconvenience.
> >
> > I posted the files (vtrace_05_no_order and vtrace_stop_on_9405) again.
> >
> > I can see the message:
> > *** -9405 / RETURN SEGMENT 1 (1 part, len: 96)
> > update_fc, errpos: 12, sqlstate: 'S9405'
> > errortext PART (1 argument, size: 126760)
> > buf(35): 'System error: AK Datapart too short'
> >
> > in both files. But I am not sure if the vtrace stops at the first
> > occurrence of the error.
> >
> > Hope this helps,
> >
> > Thanks for your help,
> > Jenny
> > >
> >
> > On Thu, 2002-09-26 at 08:08, Becker, Holger wrote:
> > > Hi Jenny,
> > >
> > > unfortunately I can't find any error within your files also
> > in the "stop on error" file.
> > > I'm wondering why because this feature should stop the
> > vtrace at the first occurrence of the error.
> > >
> > > Please try it again with a larger vtrace area to avoid
> > overwiting of the vtrace.
> > > You could enlarge the vtrace area by setting TRACE_PAGES_US
> > to a higher value.
> > > But keep in mind that this increase memory consumption of
> > the SAP DB kernel.
> > >
> > > Sorry for this inconvenience.
> > >
> > > Regards,
> > > Holger
> > >
> > > > -----Original Message-----
> > > > From: Jenny Zhang [mailto:[EMAIL PROTECTED]]
> > > > Sent: Mittwoch, 25. September 2002 19:59
> > > > To: Becker, Holger
> > > > Subject: RE: -9405 AK Datapart too short
> > > >
> > > >
> > > > Holger,
> > > >
> > > > I posted two vtrace files at
> > http://www.osdl.org/archive/jenny/bugrep/
> > > > yesterday. Can you let me know if the data is helpful,
> > or do you need
> > > > more data?
> > > >
> > > > Thanks for your help,
> > > > Jenny
> > > > On Mon, 2002-09-23 at 02:56, Becker, Holger wrote:
> > > > > Hi Jenny,
> > > > >
> > > > > the vtrace shows many lines like
> > > > >
> > > > > ...
> > > > > TOO MANY PACKETS
> > > > > UNKNOWN: 17 1
> > > > > UNKNOWN: 17 1
> > > > > ...
> > > > >
> > > > > It seems to be a bug in the order vtrace which prevents
> > > > > to find the -9405 error in the trace.
> > > > >
> > > > > We will fix this bug in the next release.
> > > > > But til then could you please provide me a default vtrace
> > > > > without order option?
> > > > > Perhaps this will bring some light into the darkness.
> > > > >
> > > > > Sorry for the inconvenience.
> > > > >
> > > > > Kind regards,
> > > > > Holger
> > > > > SAP Labs Berlin
> > > > >
> > > > >
> > > > >
> > > > > > -----Original Message-----
> > > > > > From: Jenny Zhang [mailto:[EMAIL PROTECTED]]
> > > > > > Sent: Samstag, 21. September 2002 00:27
> > > > > > To: Koetter, Thomas Theodor
> > > > > > Cc: [EMAIL PROTECTED]
> > > > > > Subject: RE: -9405 AK Datapart too short
> > > > > >
> > > > > >
> > > > > > Thomas,
> > > > > >
> > > > > > The following is the last e-mail I sent to Elke. She
> > > > complained that
> > > > > > the trace files I posted does not contain the 9405 error. So
> > > > > > I replied
> > > > > > and pointed out where the error got logged. I might have
> > > > > > confused people
> > > > > > by posting several questions/bugs.
> > > > > >
> > > > > > I think what Mark reported is a different problem. But
> > > > it seems to me
> > > > > > that this 9405 error is related to multi-thread ODBC. Since
> > > > > > it does not
> > > > > > happen when I feed the data to the database using my
> > > > > > single-thread tool.
> > > > > >
> > > > > > I posted the trace file at
> > > > http://www.osdl.org/archive/jenny/bugrep/
> > > > > > Comparing the files sqltrace_driver.log.9405 and
> > > > > > sqltrace_tool.log.9405,
> > > > > > I found that those who failed and caused the 9405 error
> > > > do not have
> > > > > > a 'PARAMETER DESCRIPTION:' following the 'CALL
> > > > BUY_REQUEST_RETURN' and
> > > > > > do not have OUTPUT.
> > > > > >
> > > > > > Regards,
> > > > > > Jenny
> > > > > > ===================
> > > > > >
> > > > > >
> > > > > > Elke,
> > > > > >
> > > > > > Thanks for your reply.
> > > > > >
> > > > > > I might have confused you by sending several trace files. I
> > > > > > checked the
> > > > > > files I posted at: http://www.osdl.org/archive/jenny/bugrep/
> > > > > > There are four files regarding to 9405 error:
> > > > > >
> > > > > > 1. knldiag.9405 is the kernel diaglog. It has 4 'ERR
> > > > 51080 SYSERROR
> > > > > > -9405 AK Datapart too short' messages in it.
> > > > > > 2. sqltrace_driver.log.9405 is the odbc trace for
> > > > > > multi-thread driver.
> > > > > > There is no error in this file. But SQLExecute returns an
> > > > > > error to the
> > > > > > application.
> > > > > > 3. sqltrace_tool.log.9405 is the odbc trace for
> > > > single-thread tool.
> > > > > > SQLExecute returns successfully.
> > > > > > 4. vtrace_9405 is the vtrace file for database kernel
> > > > when 9405 error
> > > > > > happens. In this file, 9405 is recorded:
> > > > > > ...
> > > > > > KB05 id482043/7 return_result
> > > > > > TOO MANY PACKETS
> > > > > > UNKNOWN: 17 1
> > > > > > UNKNOWN: 17 1
> > > > > > 00010000 00000000 00000060 00000000 00000001
> > > > 00010002 53393430
> > > > > > 3543DB0C 00000000 00000005 00000000 00000000
> > > > 00000006 00010000
> > > > > > 00000023 00000028 EF010053 79737465 6D206572
> > > > 726F723A 20414B20
> > > > > > 44617461 70617274 20746F6F 2073686F 7274FFFF
> > > > FFFFFF2E 2E2E2E2E
> > > > > > 2E2E2E2E 2E2E2E2E 2E2E2E2E 2E2E2E2E 2E2E2E
> > > > > >
> > > > > >
> > > >
> > '...........`............S9405C.............................#...(.'
> > > > > > '..System error: AK Datapart too
> > > > > > short............................'
> > > > > > '.'
> > > > > > >KB05 id482043/7 get direct SYS1CAT key(12): 00000000
> > > > > > 000002BC 0008010F
> > > > > > >b02get key(12): 00000000 000002BC 0008010F
> > > > > > b02get root 4; e_ok
> > > > > > ...
> > > > > >
> > > > > > The kernel version is 7.3.0.25. The ODBC driver version is
> > > > > > 7.3.0.25.
> > > > > >
> > > > > > If you need more information, please let me know.
> > > > > >
> > > > > > Regards,
> > > > > >
> > > > > > Jenny
> > > > > > > Mhm, there we have a problem, by now you sent
> > several traces and
> > > > > > whatever
> > > > > > > but none with a -9405 in. Therefore we do not see anything
> > > > > > about this
> > > > > > error
> > > > > > > and/or the reason for this. Sorry, but by now there is
> > > > no chance for
> > > > > > us
> > > > > > > to check what may have happened with your special dbproc.
> > > > > > > And on the other hand, both guys trying to find out
> > > > will be out of
> > > > > > office
> > > > > > > next (half of the) week.
> > > > > > > Please try to prepare a vtrace with -9405 in AND
> > > > not more than
> > > > > > > 5 parallel working users. Otherwise the trace-output will
> > > > > > be VERY hard
> > > > > > > to handle in the case we/you have (and had with the
> > > > trace you sent).
> > > > > > >
> > > > > > > Thank you
> > > > > > > Elke
> > > > > > > SAP Labs Berlin
> > > > > > >
> > > > > >
> > > > > > On Fri, 2002-09-20 at 05:52, Koetter, Thomas Theodor wrote:
> > > > > > > Hi Jenny
> > > > > > >
> > > > > > > Hm, I think, there is a misunderstanding.
> > > > > > > Initially I thought that the dbproc-problem, which
> > was resolved
> > > > > > > by the last ODBC driver update, was responsible for the
> > > > > > > error. Apparently this is not the case.
> > > > > > >
> > > > > > > What did you discuss with Elke? Now she is on holidays until
> > > > > > > next Tuesday, so I can't ask she.
> > > > > > >
> > > > > > > Actually, I have no idea what causes the -9405.
> > > > > > >
> > > > > > > At last, one question. Is this error related to the
> > > > > > > multithreaded problem Mark Wong reported?
> > > > > > > For sure, there is a problem in multithread use of
> > the driver
> > > > > > > which is not found yet. If the -9405 does not occur in
> > > > > > > single thread scenarios, please wait until the multithread
> > > > > > > problem is solved (which seems to be not so easy to tackle).
> > > > > > >
> > > > > > >
> > > > > > > Greetings Thomas
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > ----------------------------------------------
> > > > > > > Dr. Thomas K�tter
> > > > > > > SAP DB, SAP Labs Berlin
> > > > > > >
> > > > > > >
> > > > > > > SAP DB is open source. Get it! www.sapdb.org
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > --
> > > > > > Jenny Zhang
> > > > > > Open Source Development Lab Inc
> > > > > > 15275 SW Koll Parkway - Suite H
> > > > > > Beaverton, OR 97006
> > > > > > (503)626-2455 ext 31
> > > > > >
> > > > > > _______________________________________________
> > > > > > sapdb.general mailing list
> > > > > > [EMAIL PROTECTED]
> > > > > > http://listserv.sap.com/mailman/listinfo/sapdb.general
> > > > > >
> > > > --
> > > > Jenny Zhang
> > > > Open Source Development Lab Inc
> > > > 15275 SW Koll Parkway - Suite H
> > > > Beaverton, OR 97006
> > > > (503)626-2455 ext 31
> > > >
> > --
> > Jenny Zhang
> > Open Source Development Lab Inc
> > 15275 SW Koll Parkway - Suite H
> > Beaverton, OR 97006
> > (503)626-2455 ext 31
> >
--
Jenny Zhang
Open Source Development Lab Inc
15275 SW Koll Parkway - Suite H
Beaverton, OR 97006
(503)626-2455 ext 31
_______________________________________________
sapdb.general mailing list
[EMAIL PROTECTED]
http://listserv.sap.com/mailman/listinfo/sapdb.general
