Simpler Navigation for Servers and Operating Systems - Please Update Your Bookmarks
Completed: a much simpler Servers and Operating Systems section of the Community. We combined many of the older boards, so you won't have to click through so many levels to get at the information you need. Check the consolidated boards here as many sub-forums are now single boards.
If you have bookmarked forums or discussion boards in Servers and Operating Systems, we suggest you check and update them as needed.
General
cancel
Showing results for 
Search instead for 
Did you mean: 

Seg fault in OCI interface (Oracle 8.1.5)

SOLVED
Go to solution
Daavid Turnbull
Frequent Advisor

Seg fault in OCI interface (Oracle 8.1.5)

Dear Database dudes,

We have a problem that shows up every now an then on our systems. I tried posting about it to the Oracle forums a couple of years back but never got any meaningful response. Recently it got back on top of the pile.

We cannot reliably reproduce this problem and only see it on one in every few thousand oracle call interface calls. It's occurance is basically at random (ie it is not after x amount of calls - we might see it twice in one day and then not for 3 weeks).

Last time when I posted to the Oracle forums I had an example with two transactions making exactly the same request data wise.

I have a core file and stack trace (see below).

I know that 8.1.5 is no longer supported but we have not been able to reliably upgrade so for the time being we are stuck with this version.
________Stack Trace____________

Wildebeest was built for PA-RISC 1.1 or 2.0 (narrow), HP-UX 11.00.
..
Core was generated by `dbues'.
Program terminated with signal 11, Segmentation fault.

warning: The shared libraries were not privately mapped; setting a
breakpoint in a shared library will not work until you rerun the program.

#0 0xc01bbbf4 in memcpy () from /usr/lib/libc.2
(gdb) bt
#0 0xc01bbbf4 in memcpy () from /usr/lib/libc.2
#1 0x46ad0c in nioqrc ()
#2 0x28b3d8 in ttcdrv ()
#3 0x471994 in nioqwa ()
#4 0x204a64 in upirtrc ()
#5 0x3f7b0 in kpuexe ()
#6 0x22911c in upiex0 ()
#7 0x2288e0 in upiexn ()
#8 0x21d6f4 in ociexe ()
#9 0x3b338 in CSSG_CPR::retrieveDestinationDetails (this=0x401a2620, protocol=0x7f72f124) at cprcom.cpp:1088
#10 0x3c210 in CSSG_CPR::executeCommand (this=0x401a2620, prot=0x7f72f124) at cprcom.cpp:1544
#11 0x3c048 in CSSG_CPR::processOraMessage (this=0x401a2620, prot=0x7f72f124) at cprcom.cpp:1471
#12 0x3bf6c in CSSG_CPR::processMessage (this=0x401a2620, msg=@0x7f72f120) at cprcom.cpp:1426
#13 0x4ccbdc in CSSG_dB::processMessageMgr (args=0x7f72f120) at com.cpp:532
#14 0xc04ea0b0 in __pthread_create_system () from /usr/lib/libpthread.1
(gdb) quit
Behold the turtle for he makes not progress unless he pokes his head out.
5 REPLIES
Eric Antunes
Honored Contributor

Re: Seg fault in OCI interface (Oracle 8.1.5)

Hi,

You could always upgrade to 8.1.7.4 (this is the latest version for 8i) and see if this continue to happen...

Best Regards,

Eric Antunes
Each and every day is a good day to learn.
Daavid Turnbull
Frequent Advisor

Re: Seg fault in OCI interface (Oracle 8.1.5)

We have tried to upgrade in our lab environment we have not yet succeeded in doing it cleanly. Unless we can formulate an upgrade process that is bullet proof we won't be trying it in production.

What I was hoping was that somebody knew of a patch affecting this area. The transaction it is failing on is very simple and basic and only involves a few hundred bytes of data.
Behold the turtle for he makes not progress unless he pokes his head out.
Eric Antunes
Honored Contributor
Solution

Re: Seg fault in OCI interface (Oracle 8.1.5)

There is one patch about "Segmentation": patch 1733170 ("RDBMS Server: Patch
SEGMENTATION VIOLATION IN CLIENT IN GETORACLETYPEADT AGAINST 8.0X/8I SERVER"). Another patch - refers interfaces - is patch 1949094 ("RDBMS Server: Patch
HOA_TREAT_AS_GRAPHIC FLAG NOT BEING KEPT IN HOADAMOD ACROSS INTERFACE"). Both patches upgrade your RDBMS to 8.1.7. Finally, about core dumps you have all those patches:

Patch Description Release Updated Size
2514869 RDBMS Server: Patch
SELECT STATEMENT END UP IN CORE DUMP WHEN CURSOR_SHARING=FORCE IS SET 8.1.7.4 27-JUL-2004 14K
2288481 RDBMS Server: Patch
SELECT DATA FROM A PARTITIONED TABLE CAUSES COREDUMP, ORA-600[15212] OR ORA-7445 8.1.7.4 18-MAR-2004 49K
2523782 RDBMS Server: Patch
QUERY REWRITE COREDUMPS WITH LARGE NUMBER OF MATERIALIZED VIEWS 8.1.7.4 14-OCT-2003 55K
1390304 RDBMS Server: Patch
CORE LDXGDT FUNCTION IS NOT THREAD SAFE.. 8.1.7.0 29-OCT-2002 8.3K
2052836 RDBMS Server: Patch
CORE DUMP WITH CURSOR_SHARING=FORCE AND USING THE CALL OPERAND 8.1.7.3 26-JUN-2002 20K
1194937 Oracle Database Family: Patch
8.1.6.0:SQLPLUS core with 816 on HP 8.1.6.0 08-JUN-2002 3.3K
1748759 RDBMS Server: Patch
A CORE DUMP OCCURD WHEN INCREASES A COUNT OF PREFETCH. 8.1.7.3 15-APR-2002 11K
2220597 RDBMS Server: Patch
ORA-7445 AND CORE DUMPS TRYING TO CONNECT TO OFFLINE DATABASE IN /RDBMS/LOG 8.1.7.3 21-FEB-2002 33K
1247328 RDBMS Server: Patch
CORE DUMP IN KPCDRSZ().

Best Regards,

Eric Antunes
Each and every day is a good day to learn.
Eric Antunes
Honored Contributor

Re: Seg fault in OCI interface (Oracle 8.1.5)

To migrate to 8.1.7.4 you have to do several sequential steps:

1 - install the 8.1.7.3 CD (doing all that is said in the README.html)
2 - install 8.1.7.4 patch
3 - install RDBMS patches 2073948, 2225569, 2346568, 2386334 and 2401670.
4 - Some initTST.ora parameters are obsoleted (SEQUENCE_CACHE_ENTRIES and SEQUENCE_CACHE_HASH_BUCKETS)
...
Each and every day is a good day to learn.
Daavid Turnbull
Frequent Advisor

Re: Seg fault in OCI interface (Oracle 8.1.5)

These are the sort of points I was after - thank-you.

I now have to go away and try them out and report the findings when I have them.
Behold the turtle for he makes not progress unless he pokes his head out.