| core with Oracle 8.1.7 on AIX 4.3.3 [message #97187] |
Thu, 10 January 2002 04:14  |
Luc Vanel
Messages: 1
Registered: January 2002
|
Junior Member |
|
|
Hello,
I've got a core in a multi-threaded program using Oracle Server 8.1.7
with Pro*C on AIX 4.3.3 / RS6000 (with 2 processors). The stack is shown
below :
>>
Segmentation fault in waitpid at 0xd01a3a28
0xd01a3a28 (waitpid+0xd0) 93a30000 stw r29,0x0(r3)
(dbx) where
waitpid(??, ??, ??) at 0xd01a3a28
sntpreap(??, ??) at 0xd10d86a0
sslsshandler(0x14) at 0xd0e42784
vp._vp_sleep(??, ??) at 0xd0012a50
evt._usched_dispatch(??, ??, ??) at 0xd0011130
<<
The program works fine on another RS6000 (single processor) with Oracle
8.1.6.
Would anybody have an idea for this ?
Thanks alot.
|
|
|
|
| Re: core with Oracle 8.1.7 on AIX 4.3.3 [message #97422 is a reply to message #97187] |
Mon, 29 April 2002 10:29  |
Frank Baratta
Messages: 1
Registered: April 2002
|
Junior Member |
|
|
Luc
We've just did an AIX 4.3.3 upgrade to 2001d+ with ORACLE 8.1.6 to 8.1.7. We are getting seg faults with waitpid on every core dump ( exactly the same as you received down to the last command - evt._usched_dispatch. The problem went away when we rolled back to AIX 4.3.3 2001a. We are still using ORACLE 8.1.7. However; we are using a database interface software called RogueWave.
Did IBM ever give you a fix for this problem. They tried to tell us that it was in our code but we do not use 'wait'. That gets called on a very low level AIX call. Were you able to resolve yourself??? I see many manual fixes to this on the web for many versions of UNIX.
Frank
|
|
|
|
|
|
] 
Blog
Search
Help
Members
Register
Login
Home
