Re: Oracle screws up Sun host

From: Steve <noone_at_nowhere.com>
Date: Wed, 23 Apr 2003 16:13:04 GMT
Message-ID: <jYypa.582947$3D1.323810@sccrnsc01>

Rick,

I am supporting a client running Oracle 8.1.7.4 on Solaris. I also feel your pain with regards to Oracle Support.

First, if you want to easily identify which ipc resources are being used by a particular instance with an 8.1.7 database, run the $ORACLE_HOME/bin/sysresv utility. After you've made all attempts to shutdown gracefully, kill processes, etc., you can manually remove the remaining shared memory segments, message queues, and semaphores for a specific instance identified by this utility with the ipcrm command.

Second, I had quite a few ORA-600 and ORA-7445 errors with 8.1.7 on Solaris until I upgraded to Oracle patchset #4 (i.e. 8.1.7.4).

These errors had plagued our 8.1.7 environments but have not occurred in our 8.1.7 production databases since I applied patchset 8.1.7.4 about 5 months ago:

ORA-00600: internal error code, arguments: [729]
ORA-00600: internal error code, arguments: [4454]
ORA-00600: internal error code, arguments: [12333]
ORA-00600: internal error code, arguments: [15851]
ORA-00600: internal error code, arguments: [17090]
ORA-00600: internal error code, arguments: [17112]
ORA-00600: internal error code, arguments: [17114]
ORA-00600: internal error code, arguments: [17116]
ORA-00600: internal error code, arguments: [17182]
ORA-00600: internal error code, arguments: [17285]
ORA-00600: internal error code, arguments: [KGHFRE2]
ORA-00600: internal error code, arguments: [kghuclientasp_03]


ORA-07445: exception encountered: core dump [kghalf()+188]
ORA-07445: exception encountered: core dump [kghalf()+348]
ORA-07445: exception encountered: core dump [kghalf()+512]
ORA-07445: exception encountered: core dump [kghalf()+700]
ORA-07445: exception encountered: core dump [kghfrempty()+84]
ORA-07445: exception encountered: core dump [kghfrf()+356]
ORA-07445: exception encountered: core dump [kghfrmrg()+296]
ORA-07445: exception encountered: core dump [kghfrmrg()+56]
ORA-07445: exception encountered: core dump [kghfrmrg()+88]
ORA-07445: exception encountered: core dump [kghidmp()+392]
ORA-07445: exception encountered: core dump [kghrsp()+196]
ORA-07445: exception encountered: core dump [kghrst()+700]
ORA-07445: exception encountered: core dump [kghsrch()+56]
ORA-07445: exception encountered: core dump [kghtshrt()+80]
ORA-07445: exception encountered: core dump [kghuclientasp()+16]
ORA-07445: exception encountered: core dump [kgidmp()+1168]
ORA-07445: exception encountered: core dump [kkeisj()+48]
ORA-07445: exception encountered: core dump [kketac()+560]
ORA-07445: exception encountered: core dump [kkocxj()+592]
ORA-07445: exception encountered: core dump [kkocxj()+612]
ORA-07445: exception encountered: core dump [kkofmx()+1324]
ORA-07445: exception encountered: core dump [kkofmx()+88]
ORA-07445: exception encountered: core dump [kkojnp()+5848]
ORA-07445: exception encountered: core dump [kkoorpr()+252]
ORA-07445: exception encountered: core dump [kkoppar()+168]
ORA-07445: exception encountered: core dump [kopecpy()+2692]
ORA-07445: exception encountered: core dump [lxhcsn()+16]
ORA-07445: exception encountered: core dump [memcpy()+28]
ORA-07445: exception encountered: core dump [memcpy()+76]
ORA-07445: exception encountered: core dump [memset()+56]
ORA-07445: exception encountered: core dump [opiomc()+172]
ORA-07445: exception encountered: core dump [opiosq0()+860]
ORA-07445: exception encountered: core dump [pfrdprintf()+52]

ORA-07445: exception encountered: core dump [pfripDr_purge_DL01_] ORA-07445: exception encountered: core dump [ssdgetcall()+208]

Steve

"Rick Denoire" <100.17706_at_germanynet.de> wrote in message news:ldjaavstv22btbdab86jjl8ts2fqc6ccm2_at_4ax.com...
> empete2000_at_yahoo.com (Pete's) wrote:
>
> >
> >The reason for having to "kick the box" since you are killing
> >processes at the command line, is probably due to the memory segments
> >not being cleaned up. Check out metalinks, this will tell you how to
> >clean up memory segments. You may not think Oracle Support is the
> >best, could it be better, yes. Most of the time, you'll be getting
> >someone who can help you work the problem or find someone that can.
>
> I appreciate your hint about memory segments.
>
> Bye
> Rick Denoire
Received on Wed Apr 23 2003 - 11:13:04 CDT