Skip to Main Content
  • Questions
  • dump files created by internal error

Breadcrumb

May 4th

Question and Answer

Tom Kyte

Thanks for the question, Khalid.

Asked: October 10, 2004 - 10:14 am UTC

Last updated: October 18, 2012 - 8:19 am UTC

Version: 9.2.0.5

Viewed 10K+ times! This question is

You Asked

Dear Tom.

Last month I have alot of internal erro ora-00600 that generates dump files. I want to know how to interpret them. I need to know the sections of the dump for example stack dump and processs. what was the users that were login at the that time from witch machin and what is the application that was running. any thing that might help me.

Regards,
Khalid.

and Tom said...

you do not interpret them, development uses them to determine the cause of the internal (unexpected) error.

Please contact support -- ora-600

[tkyte@tkyte-pc-isdn tkyte]$ oerr ora 600
00600, 00000, "internal error code, arguments: [%s], [%s], [%s], [%s], [%s], [%s], [%s], [%s]"
// *Cause: This is the generic internal error number for Oracle program
// exceptions. This indicates that a process has encountered an
// exceptional condition.
// *Action: Report as a bug - the first argument is the internal error number

[tkyte@tkyte-pc-isdn tkyte]$


the error message is very clear on this!

Just publishing this so everyone can see "ora-600" -> please contact support (so does ora-3113 typically and ora-7445)

Rating

  (23 ratings)

Is this answer out of date? If it is, please let us know via a Comment

Comments

Disappointed

Khalid, October 10, 2004 - 11:55 am UTC

Dear Tom.

I know this i just want to know how can i get more info from those dumps.

for example if you want to know what is the sql statement that raise th error.

Regards,
Khalid.

Tom Kyte
October 10, 2004 - 12:31 pm UTC

sorry if you are "disappointed" but beyond the obvious information in there (stuff that needs no explanation) there is *nothing in there* really FOR US.

It is a dump
It is for the developers
To figure out "where it all went wrong"

if you contact support
AND
if you give them the dumps
THEN
your level of disappointment will most likely decrease as they will use it to query the problem database, find related events/occurrences, diagnose if you have the same issue and see if there is not already a fix in place -- or if you have discovered something "new"


You cannot really "get more info" from it -- other then what you can read and "obviously make sense of" (like "the current sql for these session when we crashed was ...." and such)

it is just a trace file, it is designed for support -- not for us.

Bob Watkins, October 10, 2004 - 2:43 pm UTC

While I agree that the purpose of an ORA-600 dump is to help Oracle's developers, I have also found it useful to do a MetaLink search on the string "ORA-600 [nnnnn]" where nnnnn is the contents of the first set of brackets in the message. (Note that "ORA-600" often works better than "ORA-00600" but try both.)

Often the particular reason code has been seen before and a solution or workaround developed. Before calling support, I can sometimes solve my problem by seeing if there's a direct match for the same platform and version in MetaLink.

Tom Kyte
October 10, 2004 - 6:34 pm UTC

that would be the "obvious part" no?

at the top of the file, what do we see -- things like

a) "here is the error message"
b) "here is the sql"
c) "here is the call stack"

A and C are useful for searching on metalink (that's what support will do as well).

B is useful to find out where it happened and maybe work around it by doing something different.

sorry -- guess I could have said that, it just seemed "obvious"




Satisfied

Khalid, October 11, 2004 - 2:39 am UTC

Thank you.

To the original poster:

Dan Kefford, October 15, 2004 - 12:05 pm UTC

I recently found these MetaLink articles quite useful:

How to deal with ORA-600 Internal Errors
</code> http://metalink.oracle.com/metalink/plsql/showdoc?db=Not&id=144652.1

How to find the offending SQL from a trace file
http://metalink.oracle.com/metalink/plsql/showdoc?db=NOT&id=154170.1%20&blackframe=0

Troubleshoot an ORA-600 Error Using the ORA-600 Argument Lookup Tool 
http://metalink.oracle.com/metalink/plsql/showdoc?db=NOT&id=153788.1 <code>

As for determining the user, machine, and application, grep the trace file for, 'user:', 'machine:', and 'application name:'.


600 Error

Mark, August 26, 2005 - 1:27 pm UTC

Hi Tom,

REcently had this event and this is how I dealt with it. Your feedback is appreciated.

Database spun out and hung. Checked the ALERT log and discovered this:

Fri Aug 26 11:37:43 2005
Errors in file C:\Oracle\admin\PHYDB001\udump\ORA05600.TRC:

Fri Aug 26 11:37:50 2005
ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE
Fri Aug 26 11:37:50 2005
ORA-1636 signalled during: ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE...
Fri Aug 26 11:37:54 2005
ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE
Fri Aug 26 11:37:54 2005
ORA-1636 signalled during: ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE...
Fri Aug 26 11:38:56 2005
ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE
Fri Aug 26 11:38:56 2005
ORA-1636 signalled during: ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE...
Fri Aug 26 11:38:59 2005
Errors in file C:\Oracle\admin\PHYDB001\udump\ORA04772.TRC:
ORA-00600: internal error code, arguments: [kcbget_24], [1], [11], [1090539590], [36081], [219538], [0], []

Fri Aug 26 11:39:08 2005
ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE
Fri Aug 26 11:39:08 2005
ORA-1636 signalled during: ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE...
Fri Aug 26 11:39:22 2005
Thread 1 advanced to log sequence 129329
Current log# 4 seq# 129329 mem# 0: E:\PHYDB001\LOG4A.ORA
Current log# 4 seq# 129329 mem# 1: E:\PHYDB001_B\LOG4B.ORA
Fri Aug 26 11:39:22 2005
ARC2: Beginning to archive log# 9 seq# 129328
Fri Aug 26 11:39:48 2005
ARC2: Completed archiving log# 9 seq# 129328
Fri Aug 26 11:40:51 2005
Error 600 trapped in 2PC on transaction 33.89.51207. Cleaning up.
Error stack returned to user:
ORA-00600: internal error code, arguments: [kcbget_24], [1], [11], [1090539590], [36081], [219538], [0], []
Fri Aug 26 11:40:52 2005
Error 2051 trapped in 2PC on transaction 33.89.51207. Cleaning up.
Fri Aug 26 11:40:52 2005
Error 2051 trapped in 2PC on transaction 33.89.51207. Cleaning up.
Fri Aug 26 11:40:52 2005
Error stack returned to user:
ORA-02051: another session in same transaction failed
Fri Aug 26 11:40:52 2005
Error stack returned to user:
ORA-02051: another session in same transaction failed
***
Corrupt block relative dba: 0x5141abb6 (file 325, block 109494)
Bad header found during preparing block for write
Data in bad block -
type: 6 format: 2 rdba: 0x41004ff9
last change scn: 0x0000.08c4ac71 seq: 0x1 flg: 0x00
consistency value in tail: 0xac710601
check value in block header: 0x0, block checksum disabled
spare1: 0x0, spare2: 0x0, spare3: 0x0
***
Fri Aug 26 11:40:55 2005
Errors in file c:\oracle\ora81\RDBMS\trace\phydb001DBW0.TRC:
ORA-00600: internal error code, arguments: [kcbzpb_1], [1363258294], [4], [1], [], [], [], []

Fri Aug 26 11:41:00 2005
ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE
Fri Aug 26 11:41:00 2005
ORA-1636 signalled during: ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE...
Fri Aug 26 11:41:04 2005
ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE
Fri Aug 26 11:41:04 2005
ORA-1636 signalled during: ALTER ROLLBACK SEGMENT IMPORT_RBS_VLARGE ONLINE...
Fri Aug 26 11:41:45 2005
Errors in file c:\oracle\ora81\RDBMS\trace\phydb001DBW0.TRC:
ORA-00600: internal error code, arguments: [kcbzpb_1], [1363258294], [4], [1], [], [], [], []

Fri Aug 26 11:41:45 2005
DBW0: terminating instance due to error 600

Looking into this, I discover this:
12:46:03 HT4@PHYDB001-LIVE> ed
Wrote file afiedt.buf

1 SELECT segment_name, partition_name
2 FROM dba_extents
3 WHERE 109494 BETWEEN block_id AND (block_id + blocks - 1)
4* AND file_id = 325
12:46:07 HT4@PHYDB001-LIVE> /
SEGMENT_NAME
------------------------------------------------------------------------------
PARTITION_NAME
------------------------------
PP_C_G_P_P_B_P_P_T_E_S_A

And Trc file info:
*** 2005-08-26 11:37:43.804
*** SESSION ID:(44.48355) 2005-08-26 11:37:43.788
oer 8102.2 - obj# 219538, rdba: 0x5141abb6
kdk key 8102.2:
ncol: 12, len: 55
key: (55):
...

AND

*** SESSION ID:(309.16146) 2005-08-26 11:38:59.288
*** 2005-08-26 11:38:59.288
ksedmp: internal or fatal error
ORA-00600: internal error code, arguments: [kcbget_24], [1], [11], [1090539590], [36081], [219538], [0], []
...

With SQL statements hitting off the PAT_PROCEDURE table in question....

12:46:32 HT4@PHYDB001-LIVE> select object_name from user_objects where object_id = 219538 /* from trc files */;
OBJECT_NAME
------------------------------------------------------------------------------
PP_C_G_P_P_B_P_P_T_E_S_A

And

12:47:27 HT4@PHYDB001-LIVE> select table_name from user_indexes where index_name = 'PP_C_G_P_P_B_P_P_T_E_S_A';
TABLE_NAME
------------------------------
PAT_PROCEDURE

And

13:10:12 HT4@PHYDB001-LIVE> show parameter db_block_check
NAME_COL_PLUS_SHOW_PARAM TYPE VALUE_COL_PLUS_SHOW_PARAM
---------------------------------------------------------------- ------- -------------------------
db_block_checking boolean FALSE
db_block_checksum boolean TRUE

My recommendation was to drop and recreate the offending index.

Any comments? And does db_block_checking = FALSE a bad idea? I had read on this site that there may be performance implications setting this = TRUE.

Great as always,
Mark

Tom Kyte
August 26, 2005 - 2:00 pm UTC

corrupt index block -- drop + recreate can be the fastest way to fix it, yes.

There are performance implications to everything -- timed_statistics for example. But I would *never* turn that one off. If you are having repeated corruptions, I'd turn on all of the checking features to catch it as soon as possible - better to diagnose why they are happening

ora-00600 internal error

piyush, September 16, 2005 - 1:20 pm UTC

I work on oracle9i production server. version of oracle is 9.2.0.4.0. and i got this kind of errors. Is there any bug which i have to fix thru patches??? plz give me the proper guidence. here is the error...
ORA-00600: internal error code, arguments: [17285], [0xFFFFFFFF7C5F0CF0], [1], [0x3C92BD680], [], [], [], []
ORA-07445: exception encountered: core dump [0000000102153108] [SIGSEGV] [Address not mapped to object] [0x000000018] [] []
ORA-00600: internal error code, arguments: [17285], [0xFFFFFFFF7C5F0CF0], [1], [0x3C92BD680], [], [], [], []
ORA-07445: exception encountered: core dump [0000000102153108] [SIGSEGV] [Address not mapped to object] [0x000000018] [] []
ORA-07445: exception encountered: core dump [0000000102153108] [SIGSEGV] [Address not mapped to object] [0x000000018] [] []
ORA-00600: internal error code, arguments: [17285], [0xFFFFFFFF7C5F0CF0], [1], [0x3C92BD680], [], [], [], []


Tom Kyte
September 16, 2005 - 2:01 pm UTC

You need to work issues like this via support.

Boring

boring, September 30, 2005 - 8:26 am UTC

Boring

Tom Kyte
September 30, 2005 - 10:06 am UTC

exciting...

but thanks for playing!

ORA-00600

Laxman Kondal, December 13, 2005 - 8:42 am UTC

Hi Tom

While executing this query got ORA-0060 and when reframed same query in where clause 'NOT IN' to just 'IN' and add '<>' to get same result it worked fine;

ops@OPSANDD> SELECT c.*
2 FROM (SELECT s.sys_id,
3 (SELECT s.NAME || ' ' || model_version FROM SYSTEMS sx WHERE sx.sys_id = s.sys_id) AS sys_name,
4 'AV-8B' AS ac_name,
5 (SELECT aii_id FROM AII WHERE NAME = 'AV-8B' AND mission = 'POAI') AS aii_id
6 FROM SYSTEMS s
7 WHERE parent_sys_id = 42
8 AND s.sys_id = 2884
9 AND rollup_sys_id IS NULL
10 AND active_ind = 'Y' ) c
11 WHERE c.aii_id NOT IN (SELECT aii_id FROM MAINT_REF);
WHERE c.aii_id NOT IN (SELECT aii_id FROM MAINT_REF)
*
ERROR at line 11:
ORA-00600: internal error code, arguments: [qctcte1], [0], [], [], [], [], [], []
-------------------------------------------------------
and then I changed the where clause, last line

ops@OPSANDD> SELECT c.*
2 FROM (SELECT s.sys_id,
3 (SELECT s.NAME || ' ' || model_version FROM SYSTEMS sx WHERE sx.sys_id = s.sys_id) AS sys_name,
4 'AV-8B' AS ac_name,
5 (SELECT aii_id FROM AII WHERE NAME = 'AV-8B' AND mission = 'POAI') AS aii_id
6 FROM SYSTEMS s
7 WHERE parent_sys_id = 42
8 AND s.sys_id = 2884
9 AND rollup_sys_id IS NULL
10 AND active_ind = 'Y' ) c
11 --WHERE c.aii_id NOT IN (SELECT aii_id FROM MAINT_REF)
12 WHERE c.aii_id IN (SELECT aii_id FROM MAINT_REF WHERE aii_id <> c.aii_id);

no rows selected

ops@OPSANDD>
--------------------------------------------------

Could you please say what makes this query to get ORA-00600 with 'NOT IN'?

Thanks for this greate source for Oracle education and knowledge.

Thanks and regards.

Tom Kyte
December 13, 2005 - 9:47 am UTC

ora-600 = contact support, you've hit a "product issue"

ORA-00600

Laxman Kondal, December 13, 2005 - 11:52 am UTC

Hi Tom

Thanks for reply and just curious whatÂ’s "product issue".

It's Oracle9i R2 clean install and schema was created using Designer9i.

Thanks and regards.

Tom Kyte
December 13, 2005 - 1:21 pm UTC

It is a euphemism for "bug" of course :)

ORA-07445: exception encountered: core dump

Clear, July 03, 2006 - 12:56 pm UTC

Hi Tom,

I got an error from my alter.log file as :
Thread 1 advanced to log sequence 384751 Current log# 6 seq# 384751 mem# 0: /home/hapi/ORACLE/OraHome1/oradata/aps9/redolog_6a.ora
Current log# 6 seq# 384751 mem# 1: /home/db80/oradata/aps9/redolog_6b.ora
Sun Jul 2 21:00:01 2006
Errors in file /home/hapi/ORACLE/OraHome1/admin/aps9/udump/aps9_ora_10765.trc:
RRA-07445: exception encountered: core dump [kziasfc()+160] [SIGSEGV] [Address not mapped to object] [0] [] []
RRA-01017: invalid username/password; logon denied

The following is from home/hapi/ORACLE/OraHome1/admin/aps9/udump/aps9_ora_10765.trc:
Unix process pid: 10765, image: oracle@hapi (TNS V1-V3)

*** SESSION ID:(21.2843) 2006-07-02 21:00:00.847
Exception signal: 11 (SIGSEGV), code: 1 (Address not mapped to object), addr: 0x0, PC: kziasfc()+160
*** 2006-07-02 21:00:01.390
ksedmp: internal or fatal error
ORA-07445: exception encountered: core dump [kziasfc()+160] [SIGSEGV] [Address not mapped to object] [0] [] []
ORA-01017: invalid username/password; logon denied

How can I know what user login failed?

Thanks!



Tom Kyte
July 07, 2006 - 7:20 pm UTC

not knowing the context of this - unlikely you can. but - in any case, please contact support with those trace files.

ORA-07445: exception encountered: core dump

A reader, July 05, 2006 - 9:59 am UTC

Hi Tom,

I have to re-post my question since there is an error in my message:

I got an error from my alter.log file as :
Thread 1 advanced to log sequence 384751 Current log# 6 seq# 384751 mem# 0:
/home/hapi/ORACLE/OraHome1/oradata/aps9/redolog_6a.ora
Current log# 6 seq# 384751 mem# 1: /home/db80/oradata/aps9/redolog_6b.ora
Sun Jul 2 21:00:01 2006
Errors in file /home/hapi/ORACLE/OraHome1/admin/aps9/udump/aps9_ora_10765.trc:
ORA-07445: exception encountered: core dump [kziasfc()+160] [SIGSEGV] [Address
not mapped to object] [0] [] []
ORA-01017: invalid username/password; logon denied

The following is from
home/hapi/ORACLE/OraHome1/admin/aps9/udump/aps9_ora_10765.trc:
Unix process pid: 10765, image: oracle@hapi (TNS V1-V3)

*** SESSION ID:(21.2843) 2006-07-02 21:00:00.847
Exception signal: 11 (SIGSEGV), code: 1 (Address not mapped to object), addr:
0x0, PC: kziasfc()+160
*** 2006-07-02 21:00:01.390
ksedmp: internal or fatal error
ORA-07445: exception encountered: core dump [kziasfc()+160] [SIGSEGV] [Address
not mapped to object] [0] [] []
ORA-01017: invalid username/password; logon denied

How can I know what user login failed?

Thanks!


Tom Kyte
July 08, 2006 - 7:58 am UTC

see above.

Some tips on ORA-600

Serge Shmygelsky, September 08, 2006 - 4:53 am UTC

To whom it may concern:

Today my PC lost the power during Oracle shutdown. After I've restarted the PC, I found that it was not my lucky day as Oracle didn't start. In the alert log I found the following:
ksedmp: internal or fatal error
ORA-00600: Message 600 not found; No message file for product=RDBMS, facility=ORA; arguments: [3705] [1] [1] [2] [10]

So after scratching my head for a while I went to Metalink and after some mucking around found the note 352666.1.

I've done the steps proposed by the note (to be exact - recreated the controlfiles) and voila - my Oracle is up and running. It is better than sex, you know.

But I should tell you that this is a developer database. I doubt that I'd run it on production machine (especially considering that I'm not a DBA).

HIH.

Tom Kyte
September 08, 2006 - 4:29 pm UTC

you had other issues there serge - message not found, indicates environment problem there.

Message 600 not found

Serge Shmygelsky, September 11, 2006 - 3:40 am UTC

Yes, I saw it but nevertheless Oracle is working fine. Now I'm trying to find the reason for the message

Tom Kyte
September 11, 2006 - 9:54 am UTC

because your environment is not set.

Try this...

Invisible, November 17, 2006 - 12:51 pm UTC

I could probably fill TheDailyWTF.com for several weeks with my job, but I'd like to share this one with you...

We recently "inherited" a beaten-up old Oracle server from another site. I'm sitting at my desk one day, and our Director of IT is in the server room setting up this server. With him is the Deputy Director of IT.

This is an exact transcript of what I heard:

Director: Hey, do you know anything about Oracle?
Deputy: No. Why, what's up?
Director: Oracle won't open the database for some reason. It's saying something about recovery.
Deputy: Did you try the "force" option?
Director: No I didn't - let me try that... There we go! Now it opens... Thanks.
Deputy: No problem.

In retrospect, I should probably have run into the server room screaming NO! DON'T DO IT!!

However, it appears that might not have helped. Once those guys had gone home again, I had a look at the alert logs on the database. It seems roughly every 3 weeks or so Oracle crashes with an ORA-0600 or some message about "fractured blocks". And it's been doing that for years.

Now I have been running an Oracle database here for YEARS. And I have seen Oracle crash exactly once. (Something about being unable to allocate memory.) 1 crash in 5 years is slightly different to 1 crash every 3 weeks. (!) Micro$oft Windoze crashes every 3 weeks. Oracle just doesn't DO that!

As far as I am aware, the only way a live Oracle database can end up with fractured blocks is if somebody doesn't know how to do backup and restore properly. And if I understand that the term means, that would seem to explain why Oracle is giving me undefined error codes like 0600.

So... I'm thinking I should get RID of this database. Rapidly. (Fortunately it's almost empty - at the moment.)

Would you agree with this conclusion?


Tom Kyte
November 20, 2006 - 2:08 am UTC

without the actual error message - not sure what to say. But, if it is block corruption, and it has been there a long long time - then yes, a rebuild of the database might be the only good path forward since you don't have any good backups to recover with.

ora-00600

sam, January 25, 2007 - 1:08 pm UTC

Tom:

Is this something for support. metalink says "two buffers have been found in the buffer cache for the same DBA". DO you think it might be corrupted table because it seems to happen when a user tried to access a table. Oracle is 9.2.0.2


SQL> select stock_number,storage_code,warehouse_id from storage_shipment where stock_number='XX007'
;

STOCK_NUMBER STORAGE_CO WAREHOUSE_ID
-------------------- ---------- ------------
XX007 TR KANSAS


ERROR:
ORA-00600: internal error code, arguments: [kcbgtcr_6], [], [], [], [], [], [],
[]

Instance is crashed

K P Ratnaker, April 03, 2007 - 8:49 am UTC

Errors in file d:\nervecen\trace\bdump\nervecentrex_m000_3836.trc:
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x345D704] [] [] [] []

Tue Apr 03 17:28:19 2007
Errors in file d:\nervecen\trace\udump\nervecentrex_ora_2640.trc:
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x3470353] [] [] [] []

Tue Apr 03 17:28:20 2007
Errors in file d:\nervecen\trace\udump\nervecentrex_ora_2640.trc:
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x3470353] [] [] [] []
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x3470353] [] [] [] []

Tue Apr 03 17:28:22 2007
Errors in file d:\nervecen\trace\udump\nervecentrex_ora_2640.trc:
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x3470353] [] [] [] []
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x3470353] [] [] [] []
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x3470353] [] [] [] []

Tue Apr 03 17:28:26 2007
Errors in file d:\nervecen\trace\bdump\nervecentrex_pmon_1072.trc:
ORA-07445: exception encountered: core dump [ACCESS_VIOLATION] [0x3470353] [] [] [] []

Tue Apr 03 17:28:44 2007
ARC1: terminating instance due to error 472
Tue Apr 03 17:28:45 2007
Errors in file d:\nervecen\trace\bdump\nervecentrex_ckpt_3072.trc:
ORA-00472: PMON process terminated with error

Tue Apr 03 17:28:45 2007
Errors in file d:\nervecen\trace\bdump\nervecentrex_rvwr_3624.trc:
ORA-00472: PMON process terminated with error

Tue Apr 03 17:28:46 2007
Errors in file d:\nervecen\trace\bdump\nervecentrex_lgwr_2240.trc:
ORA-00472: PMON process terminated with error

Tue Apr 03 17:28:47 2007
Errors in file d:\nervecen\trace\bdump\nervecentrex_mman_744.trc:
ORA-00472: PMON process terminated with error

Tue Apr 03 17:28:47 2007
Errors in file d:\nervecen\trace\bdump\nervecentrex_dbw0_308.trc:
ORA-00472: PMON process terminated with error

Tue Apr 03 17:28:47 2007
Errors in file d:\nervecen\trace\bdump\nervecentrex_reco_2324.trc:
ORA-00472: PMON process terminated with error

Instance terminated by ARC1, pid = 3648
Dump file d:\nervecen\trace\bdump\alert_nervecentrex.log
Tue Apr 03 17:37:38 2007

Intance is terminated. how i can trace this problem. this is fifth time i got an error. Please suggest me.


Tom Kyte
April 04, 2007 - 9:45 am UTC

see my comments on the other location where you put this.

Ora-600 Error

Raghav, July 17, 2007 - 10:14 am UTC

Hi Tom

I am getting an ora-600[15201] [] [] [] [] everytime I do a SELECT table_name.* FROM table_name WHERE column IN (value 1, value2...).

And the trace files points to the problem to the last query which I was running.

The tracefile contains the following:
ksedmp
kdgeridv
kdsesic
kksmal0
kksmal

But, as you know, there is no problem with the query using the IN operator in WHERE clause.

In some other case, I am getting the same problem and the trace files points to the 'function based index' which was running at that time. The query is using "WITH" clause and the when approached Oracle they are also pointing out to the same.

I want to know from you that whether Ora-600 error is related to the queries we run or it is related to the oracle configuration / infrastructure so that instead of running here and there we can address / drive in right direction.

Thanks in advance
Raghav

Tom Kyte
July 17, 2007 - 1:11 pm UTC

ora-600, 7445, 3113 => support

It's just not always have to be ORA-600/7445

Uldis Pavuls, October 31, 2007 - 2:47 pm UTC

You know that better than I do, I'm sure, Tom. And a lot of us know not only in theory what to do if ORA-600/7445 happens.
Let me rephrase the question: you certainly rater have had hacked core dumps and know a lot about them - why don't you want to share your knowledge and to help us learn about it? For example, when core's dumped by oradebug? Mayhap a bit less work for Oracle Support - and on occasion saved time on emergency.
Best regards,
Uldis Pavuls

It's just not always have to be ORA-600/7445

Uldis Pavuls, October 31, 2007 - 2:53 pm UTC

You know that better than I do, I'm sure, Tom. And a lot of us know not only in theory what to do if ORA-600/7445 happens.
Let me rephrase the question: you certainly rater have had hacked core dumps and know a lot about them - why don't you want to share your knowledge and to help us learn about it? For example, when core's dumped by oradebug? Mayhap a bit less work for Oracle Support - and on occasion saved time on emergency.
Best regards,
Uldis Pavuls

fractured or logical?

jiang huang zheng, May 07, 2008 - 10:00 am UTC

Hello Tom

       on my test database,I did this in the sqlplus and results are:
   SQL> select *  from v$database_block_corruption;

     FILE#     BLOCK#     BLOCKS CORRUPTION_CHANGE# CORRUPTIO
---------- ---------- ---------- ------------------ ---------
         9      11860          1           10054630 LOGICAL
        29         13          1                  0 FRACTURED
        29         15          1                  0 FRACTURED
        38         28          1                  0 FRACTURED
        38        364          1                  0 FRACTURED


as you can see , the corrupt is categorized into logical and factured,Is the factured means physical corruption? 

    Thanks!

Tom Kyte
May 08, 2008 - 3:03 am UTC

http://docs.oracle.com/docs/cd/B19306_01/server.102/b14237/dynviews_1074.htm#REFRN30048

http://docs.oracle.com/docs/cd/B19306_01/backup.102/b14191/rcmconc1012.htm#sthref367

It is physically, logically corrupt - the head and tail do not match. It could be either or.


corruption change# = 0 indicates "media corrupt"

Media corrupt

jiang huang zheng, May 08, 2008 - 1:05 pm UTC

Hello Tom

Thank you for the prompt reply..
and one thing I am not very clear is " media corrupt" , is it referring to some thing is wrong with the physical disk block, disk block cant be read by oracle? The change# equals zero because oracle cant read it out from block?
Tom Kyte
May 12, 2008 - 9:50 am UTC

it is physically corrupt on disk, only half there, media must have failed us in some way - we asked to write 8k and only 4k seems to have made it. Block is corrupt physically on disk.

We can read it - we did read it - it is fractured, split in half as of two different points in time.

ORA-00600: internal error code, arguments: [qertbFetchByRowID], [], [], [], [], [], [], []

Jervis, December 29, 2011 - 11:06 am UTC

Dear Tom,

Recently when our database ran a job(the job is a query which execute every 3 minutes), sometimes it got "ORA-00600: internal error code, arguments: [qertbFetchByRowID], [], [], [], [], [], [], []".

I have done "analyze table/index validate structure", but it did not get any error.

The version of our database is 10.2.0.5.0 64bit.
OS level is AIX 5.3.

I serched the event on metalink, but got nothing useful, can you give me any suggestions?

Thank you!
Tom Kyte
December 29, 2011 - 11:54 am UTC

please utilize support for this. don't just search on metalink, open up an SR, this is at least a p2 issue.

# of sumkeys invalid

Ravi B, October 17, 2012 - 4:17 pm UTC

Hello Tom,

We use Pentaho BI/ETL tool for our ETL process. We create a dummy list partition when we initially create a table and keep adding list partitions before we load the data based on the 'list' value. It was working fine until we hit the error recently:

ORA-00600: internal error code, arguments: [# of sumkeys invalid], [], [], [], [], [], [], [], [], [], [], []

When i do:

select * from table_name it gives the error.

When i do:
select * from <table_name> partition <dummy_partition>, no error

When i do:

select * from <table_name> partition <real_partition_with_data>, no error

When i do:
alter table <table_name> truncate <dummy_partition>;
select * from <table_name> the error goes away. (Dummy partition is always empty)

I tried to reproduce this manually (not from Pentaho) but i couldn't.

I dont have oracle support account. Would you have any idea what could be causing this error and what does it mean by "# of sumkeys invalid"?

Thanks!
Tom Kyte
October 18, 2012 - 8:19 am UTC

I do not see anything related to list partitioning and this error after a quick search, sorry.