[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference orarep::nomahs::sql_services

Title:SQL/Services Forum
Notice:kits(3) ft info(7) QAR access (8) SPR access (10)
Moderator:SQLSRV::MAVRIS
Created:Thu Oct 13 1988
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2214
Total number of notes:8586

2131.0. "ftlexeerr and trashed sqlsrv$exe dump" by M5::JHAYTER () Thu Jan 23 1997 17:32

T.RTitleUserPersonal
Name
DateLines
2131.1Bugcheck dumps on Alpva are VMS process dumps; also try memory debugging.SQLSRV::OXBURYOracle Corporation, Rdb Desktop Group|DTN 381-2704Thu Jan 23 1997 19:1829
2131.2Is any debugging documented?chsr38.ch.oracle.com::ROHRThe Packers did it!Mon Jan 27 1997 10:2513
    Nothing to do with this note, but:
    
    You gurus in Sql Services, would you mind putting together a kind of
    technical article "How to debug Sql Services problems"? I mean with the
    SDA and ANA/PROC and the undocumented debug flags we should know? I am 
    running out of pages in my notebook putting information like in .1 down...
    Or is there an SQl Services internals support book or some such?
    
    Thanks,
    Regina
    
    
     
2131.3did we learn anything???M5::JHAYTERMon Jan 27 1997 13:1755
Hello Simon,

I had the customer try $ ana/proc <filename>/image=sys$system:sqlsrv$exe.
I don't think we got the expected results.

     Jerry,

     I ran the procedure you gave me. I hope it means something to you. Let
     me know what you think

     Thanks,
     Fred

     ----------------------------------------------------------------------


     CEDAR$ ana/proc SQLSRV$EXE.DMP/image=sys$system:sqlsrv$exe
     Condition signalled to take dump:


       Improperly handled condition, invalid condition handler procedure
     value.
         Signal arguments:   Number = 00000003
                             Name   = 0000043C
                                      7EDB8094
                                      0000001B

         Register dump:
         R0  = 0000000000000000  R1  = 000000007ED76DD0  R2  =
     0000000000000020
         R3  = 000000007EDB8000  R4  = 0000000000003FFF  R5  =
     000000007ED86C43
         R6  = 000000007ED87730  R7  = 000000007ED87684  R8  =
     000000007ED87170
         R9  = 000000007FF9C400  R10 = 000000007FF9D228  R11 =
     000000007FFBE3E0
         R12 = 0000000000000000  R13 = 000000007ED76F98  R14 =
     0000000000000000
         R15 = 000000007ED86000  R16 = 0000000001615ED0  R17 =
     0000000001615D70
         R18 = 0000000001615F00  R19 = 0000000000000000  R20 =
     0000000000000000
         R21 = 0000000000000002  R22 = 0000000000000000  R23 =
     000000007ED76000
         R24 = 0000000000000000  R25 = 0000000000000002  R26 =
     FFFFFFFFB083A2BC
         R27 = 000000007ED76DD0  R28 = FFFFFFFFB0838570  R29 =
     0000000001615D00
         SP  = 0000000001615CF0  PC  = 000000007EDB8094  PS  =
     300000000000001B
     %SYSTEM-F-OPCDEC, opcode reserved to Digital fault at PC=00000000,
     PS=0000043C
     CEDAR$


2131.4Nope, unfortunately, "Improperly handled condition" wasn't what we wanted...SQLSRV::OXBURYOracle Corporation, Rdb Desktop Group|DTN 381-2704Mon Jan 27 1997 19:1444
Hi Jerry,

Unfortunately, you're right, that wasn't the expected result. What you should
get is a DBG> prompt, at which one can do things like SHOW CALLS, EXAMIME, etc,
etc. (Though clearly a map, plus listings are needed, because there's no
debugging info in the image.) I can't explain why they're getting that error,
except to say that ANA/PROC is EXTREMELY touchy - the slightest whiff of a
problem and it'll blow up in all sorts of weird ways, this being a case in
point. A couple of things to try is to give the process ALL privs and make sure
the executor image isn't installed before issuing the ANA/PROC command. In any
event, see if you can get the .DMP file to us somehow - although we've never
tried it before, we might be able to find the stack and figure out what's going
on. FYI, what you should get is something like this:

Kntavn_1 $ ana/proc SQLSRV_TEST:[VMSAPI3]SQLSRV$EXE.DMP;1/ima=sys$system:sqlsrv$
exe
Condition signalled to take dump:

%SYSTEM-F-ACCVIO, access violation, reason mask=04, virtual address=27327978, PC
=0003AE70, PS=0000001B

         OpenVMS Alpha AXP DEBUG Version V6.1-000

%DEBUG-I-NOLOCALS, image does not contain local symbols
%DEBUG-I-NOGLOBALS, some or all global symbols not accessible
%DEBUG-I-NOUNIVERSALS, shareable image contains no universal symbols
%SYSTEM-F-ACCVIO, access violation, reason mask=04, virtual address=27327978, PC
=0003AE70, PS=0000001B
break on unhandled exception at SHARE$SQLSRV$EXE+175728

DBG> show call
 module name     routine name                     line       rel PC    abs PC
*SHARE$SQLSRV$EXE                                           00000000  0003AE70
*SHARE$SQLSRV$EXE                                           00000000  000399B8
*SHARE$SQLSRV$EXE                                           00000000  0003687C
*SHARE$SQLSRV$EXE                                           00000000  00034FB4
*SHARE$SQLSRV$EXE                                           00000000  00030D34
*SHARE$SQLSRV$EXE                                           00000000  00030588
*SHARE$SQLSRV$EXE                                           00000000  00030080
DBG>

Let us know what you find...

Si
2131.5M5::JHAYTERTue Jan 28 1997 16:286
Well the customer has sent me a lot of stuff, but isn't telling me much.
Finally got out of him that he is also getting an sqlbugck dump evertime
he has problems with sql services.  Something to do with segmented strings.

Now we know what the real problem is.