Oracle Database Discussions

Hello all,

Looking for some guidance with a 10.2.0.1 database on Windows 2003 Enterprise Edition server. I have an SR open with Oracle, but getting the usual run-around and no conclusive explanations.

A couple of weeks ago, the database started throwing the following in the alert log:
Errors in file d:\oracle\admin\hdprd\bdump\hdprd_cjq0_404.trc:
ORA-27300: OS system dependent operation:WaitForSingleObject failed with status: 0
ORA-27301: OS failure message: The operation completed successfully.
ORA-27302: failure occurred at: sssxcpttcs5

Checking the trace file shows:
*** 2011-03-22 17:37:21.391
*** SERVICE NAME:(SYS$BACKGROUND) 2011-03-22 17:36:53.890
*** SESSION ID:(327.1) 2011-03-22 17:36:53.890
Waited for process J000 to initialize for 60 seconds
*** 2011-03-22 17:37:21.391
Dumping diagnostic information for J000:
OS tid = 3620
Memory (Avail/Total): Ph:6548M/8190M, Ph+PgF:10684M/12074M, VA:994M/2047M
CPU Load: 5%
------------------- Call Stack Trace ---------------------
calling location entry point arg #1 arg #2 arg #3 arg #4
------------------------------------------- ------------------------------------------- -------- -------- -------- --------
7C82847C 00000000 00000000 00000000 00000000 00000000
77E41A89 77E64849 00000000 00000000 00000000 00000000
_slxefo+154 00000000 00000000 00000000 00000000 00000000
slxcfot+805 slxefo+0 00000000 00000000 00000000 00000000
lxlfopn+128 slxcfot+0 412dd744 412dd98c 60b2c20c 60b2c208
lxldlod+56 lxlfopn+0 60b382dc 412dd98c 00000000 00000000
_lxdlobj+79 00000000 60b382dc 412dd98c 0a4957f0 00000001
lxdgetobj+187 lxdlobj+0 00000001 00000000 0a49df34 00000000
lxplset+49 lxdgetobj+0 00000000 00000000 00000000 00000000
lxhLaToId+146 lxplset+0 00000001 0a49dd14 0a49df34 00000000
lxhenvquery+197 lxhLaToId+0 00000000 00000000 00000000 00000000
lxhLangEnv+352 lxhenvquery+0 00000000 00000000 00000000 00000000
62747F74 62765084 0a49dd14 00000000 0a49df34 0a49df34
60987803 60A35104 00000000 00000000 00000000 00000000
60986EEE 60986EFC 3b512c78 412df6b4 412dfd84 00000100
6098798C 60986ED8 3b512c78 412df6b4 412dfd84 00000100
npinli+78 nlstdgo+0 00000000 00000000 00000000 00000000
opiinit+142 npinli+0 00000001 00000000 00000000 00000000
opimaireal+194 _opiinit+0 00000000 00000000 00000000 00000000
opimai+92 opimai_real+0 00000003 412dfeec 00000000 00000000
BackgroundThreadStart@4+422 opimai+0 00000000 00000000 00000000 00000000
77E6482C 00000000 00000000 00000000 00000000 00000000
---------------- End of Call Stack Trace -----------------
Call stack acquisition performance stats:
setup time (lock acquis., memory alloc.): 0 ms
frame get time (time the target proc was suspended): 0 ms
symbol translation time: 157 ms
total time: 157 ms

The database remains functional while this is happening, but we are no longer able to establish new sessions to the database. The only remedy so far, as is the case with most Windows issues, is to reboot the server. The support analyst has recommended that we add the "/3gb" switch to the boot.ini as he feels we are hitting the 2GB limitation for a single process. My concern is that in looking at the trace file, it would appear that there's only around 1.6GB of memory consumed in total on the server (8190M - 6548M = 1642M), so there should be no single process that's coming even close to the 2GB limitation. I've been watching the oracle.exe process for the last week, and it always hovers around 700M. The total SGA size is about 600M.

Interestingly enough, we're in the process of downgrading the database from Enterprise Edition to Standard edition and on the development server, we've currently got the old 10.2.0.1 EE database and the new 10.2.0.5 SE database running concurrently. The EE version has encountered the same issue described above, while the SE version has not.

My other problem is that I work in the pharmaceutical industry and this is a heavily regulated database and I cannot make arbitrary changes to databases/servers without evidence to support the change.

Any Oracle/Windows experts out there that can shed some light on this problem?

Thanks