From: support (support@xiatecs.net)
Date: Sat Nov 16 2002 - 14:56:49 EST
Hi Holger,
Following what I found on Metalink regarding the error you getting. it seems
like a bug. I will suggest first to make sure that you environment variable
are set correctly and you are connected to the correct Database. try using
the second control file when starting up the database if you have one. At
the end if none works check some bug fixes that are related to the ora-600
internal error you are getting.
XiaTecs Support Team
support@xiatecs.net <mailto:support@xiatecs.net>
www.xiatecs.net <http://www.xiatecs.net>
PURPOSE:
This article discusses the internal error "ORA-600 [17585]", what
it means and possible actions. The information here is only applicable
to the versions listed and is provided only for guidance.
ERROR:
ORA-600 [17585][a][b][c][d][e]
VERSIONS:
versions 8.0.X to 8.1.X
DESCRIPTION:
This is a heap corruption in the large pool.
This exception can occur during direct I/O operations such as
create index in parallel.
The ORA-600 triggers when an inconsistency is detected in the
kernel generic file i/o data structures.
FUNCTIONALITY:
KERNEL DIRECT FILE I/O
IMPACT:
PROCESS FAILURE
MEMORY CORRUPTION
NON DATA CORRUPTIVE - No underlying data corruption.
SUGGESTIONS:
Consider disabling io slaves if asynch_io is present.
Reference bug 762490
<http://metalink.oracle.com/metalink/plsql/ml2_documents.showDocument?p_id=7
62490&p_database_id=BUG> fixed in release 8.0.6 and higher.
Problem Description:
====================
The following messages appear at regular intervals in the Alert.log file,
without any accompanying errors spooled:
Thu May 17 18:50:54 2001
Timed out trying to start shared server I101.
Thu May 17 18:53:00 2001
Timed out trying to start shared server I101.
Thu May 17 18:55:07 2001
Timed out trying to start shared server I101.
Solution Description:
=====================
The following event can assist in suppressing these messages, which can
happen
because of an underlying operating system IO slave problem:
event = "10281 trace name context forever, level 300"
This event should be set in the database init.ora, and the database should
then
be restarted.
Explanation:
============
The above message can be displayed if IO slaves creation timeout occurs.
Oracle
by default waits 120 seconds for an IO slave to 'kick off'. If this
threshold
of waiting is reached, the above messages will be spooled in the alert.log.
The
wait in seconds for IO slave creation can be extended to a larger value in
seconds to suppress these messages, but this will not resolve the problem
that
is preventing the IO slaves from starting up. Spooling of new IO slaves
happens
at the operating system level and should be addressed with your specific
operating system vendor.
References:
===========
[BUG:762490]
<http://metalink.oracle.com/metalink/plsql/ml2_documents.showDocument?p_id=7
62490&p_database_id=BUG> ORA-7445 CORE DUMP [11]. ORA-600 [17585], ORA-600
[17586]
Additional Search Words
-----------------------
-----Original Message-----
From: IBM AIX Discussion List [mailto:aix-l@Princeton.EDU]On Behalf Of
Holger.VanKoll@SWISSCOM.COM
Sent: Saturday, November 16, 2002 8:59 AM
To: aix-l@Princeton.EDU
Subject: oracle (?) problem
Hello,
I
1) shutdown an oracle-db
2) backed up all files
3) recreate fs (striped now)
4) restored everything.
There a 2 db there, one works, on the other I get:
SQL*Plus: Release 8.0.5.0.0 - Production on Fri Nov 15 21:6:42 2002
(c) Copyright 1998 Oracle Corporation. All rights reserved.
Enter user-name: trp@pfmsr.swisscom.com
Enter password:
ERROR:
ORA-01034: ORACLE not available
ORA-27101: shared memory realm does not exist
IBM AIX RISC System/6000 Error: 2: No such file or directory
How can I tell which "file or directory" does not exist? How can I find out?
Tracing the listener?
tnsping works, the listener itself is up.
I checked permissions on every file, no problem there.
ipcs also shows no segments after shutting down the db.
In oracle-log I see
ORA-00600: internal error code, arguments: [17585], [33554445], [271162576],
[], [], [], [], []
very meaningful.
Any hint?
Regards,
Holger
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 22:16:21 EDT