咨询微信: dbservice1234 7 x 24 在线支持!

【Oracleデータリカバリ】ORA-00600[2662]エラ解析

【Oracleデータリカバリ】ORA-00600[2662]エラ解析

プロのOracle Databaseの復旧サービスを提供
携帯番号: +86 13764045638 メール:service@parnassusdata.com

 

 

 

 

ORA-600[2662]エラの原因はORACLEがブロックを含んでいるSCN番号はシステムに既存したSCNより大きいである。一般的に、ロールフォワードプロセスプロセスで利用されたredoは不完全で、これはシステムSCNがディスクに書き込まれたSCNより小さいくなる。SCN adjust scn 《人工的にSCN番号を増やす方法:How to increase System Change Number by manual》

 

ORA-00600[2662]関連する bugは以下の通り:

Bug 4453449
Abstract: Flashback to guaranteed restore point in orphan inc may result in ORA-600[3020]
Versions affected: 10.2.0.1
Fixed in versions: 10.2.0.2 & 11.0
Backportable: Yes
Symptoms:
The symptom of this bug include ORA-600[3020], ORA-600[2662] after flashback
database and ORA-600[flashback_validation] during flashback database.
There may also be other symptoms.
Details:
ORA-600[3020] / ORA-600 [2662] / ORA-600 [flashback_validation] can occur
after/during multiple flashback/recovery through multiple database resetlogs
without opening the database. There may also be other symptoms which appear as
recovery related corruption errors.
Workaround:
1. If you flashback a crashed primary database, follow flashback database with open
resetlogs. Alternatively, if you’d like to completely undo flashback database,
follow flashback database with recover database without shutting down the
instance first.
2. Restore backup and recover.
Patch details:
Currently there is no one-off patch available for any platform and versions.
Bug 2899477 (Unpublished)
Abstract:ORA-600[2662] CAUSES INSTANCE CRASH
Versions affected: 9.2.0.4
Fixed in versions: 9.2.0.4 & 10.1
Backportable: Yes
Symptoms:
When you have a corrupted SCN and if the corruption is found in selexe,
getting uninitialized selenv from opiexe, then this may be the bug.
Details:
It is possible for an uninitialized variable to be passed
on for a select statement which could result in a false
ORA-600 [2662] error.
Workaround:
None
Patch details:
One-off patch available for few platforms on top of 9.2.0.4
Check the Metalink for Patch 2899477 availability.
Bug 2764106
Abstract: ORA-600 [2662] BRINGS THE DATABASE DOWN
Versions affected: 8.1.7.4 & 9.2.0.4
Fixed in versions: 9.2.0.5 & 10.1
Backportable: Yes
Symptoms:
OERI(2662) even The dependent scn present in the disk blocks are fine.
Details:
A false ORA-600 [2662] error can occur on SELECT operations
which can result in an instance crash even though there is no
underlying problem with the on disk SCN.
Workaround:
None
Patch details:
One-off patch available for few platforms on top of 8.1.7.4 & 9.2.0.4
Check the Metalink for Patch 2764106 availability.
Bug 2216823 (Unpublished)
Abstract:OERI(2662) REPORTED WHEN REUSING TEMPFILE WITH RESTORED DB
Versions affected: 9.2.0
Fixed in versions: 10.1.0
Backportable: No
Symptoms:
eg:
1. Create a TEMP tablespace.
2. Shutdown a database.
3. Copy control file, data files, and log files to another directory
(but not tempfile).
4. Restart a database.
5. Create a temporary table and insert into it, thereby causing tempfile
to be updated.
6. Shutdown a database.
7. Restore a database.
8. Restart a database.
9. Create a temporary table and insert into it.
10. Commit
^- ORA-600 [2662]
Details:
ORA-600 [2662] can occur when reusing a TEMPFILE with
a restored database.
Workaround:
The workaround is not to use the pre-existing tempfile.
Instead either backup the tempfile with rest of the database
or remove the tempfile then recreate a new tempfile once the
database is open.
Patch details:
Currently there is no one-off patch available for any platforms and versions

Bug 2054025 (Unpublished)

Abstract:ORA-600 [2662] RELATED TO KDIT.C
Versions affected: 9.0.1.2
Fixed in versions: 9.0.1.3 9.2.0.1
Backportable: No
Symptoms:
OERI:2662 possible on new TEMPORARY index block
Details:
ORA-600 [2662] possible on new TEMPORARY index block
Workaround:
None
Patch details:
Currently there is no one-off patch available for any platforms and versions
Bug 851959
Abstract : ORA-600 [2662] OCCURRED DURING CREATE SNAPSHOT AT MASTER SITE
Details :
It is possible to get ORA-600 [2662] caused by mis-adjustment of the Oracle7 SCN (in PARALLEL SERVER mode) when an Oracle8 instance selects from
it over a DBLINK
Version affected : 7.3.4.X
Fixed in version: 7.3.4.5
Workaround :
None
Patch details :
Currently there is no one-off patch available for any versions/platforms.
Bug 647927 (Unpublished)
Abstract : LOCK PROCESS DIES WITH ORA-600 [2662], [0], [40057943], [0], [40063994]
Version affected 8.0.4.X
Fixed in version : 8.0.4.2 8.0.5.0
Symptoms :
Digital Unix ONLY: OERI:2662 could occur under heavy load
Workaround :
None
Patch details :
Currently there is no one-off patch available for any versions/platforms.
Bug 5612217 (Unpublished)
Abstract : ORA-7445 [KDKBIN] LEADING TO ORA-600 [2662] DUE TO BUFFER CORRUPTION
Version affected : 9.2.0.X
Workaround :
None
Patch details :
One-off patch available for few platforms on top of 9.2.0.7
Check the Metalink for Patch 5612217 availability.
Bug 4599505 (Unpublished)
Abstract : ORA-600 [2662] error
Version affected : 10.2.0.X
Fixed in version : 11.0
Symptoms :
ORA-600[2662] after flashback database.
Workaround :
This problem may disappear by itself after the database has been opened for a while and its SCN has passed the SCN of the problematic block. This is
however not a guaranteed workaround
Patch details :
Currently there is no one-off patch available for any versions/platforms.
Bug 2998110
Abstract :ORA-600 [2662] LARGE QUERIES ON STANDBY WITH LOCALLY MANAGED TMP TBLSP
Version affected : 9.2.0.X 10.1.0.X
Fixed in version : 10.2
Symptoms :
The scn of the tempfiles is advanced but not on any other files
when the database is opened in read only mode.
Workaround :
1) Increase the sort_area_size to avoid sort on disk thus avoiding the use of the tempfiles
–OR–
2) After opening the database read only and BEFORE executing any queries
against the standby database, drop and recreate the tempfiles.
–OR–
3) If you are on 10.1 release you can set the following parameter:
_init_tempfile_on_open=TRUE
in your init.ora/spfile and bounce the database.
Setting this parameter will clear all tempfile bitmaps when the database is opened
so the database open may be take a little longer.
Patch details :
Currently there is no one-off patch available for any versions/platforms.
This bug is fixed in 10.2 and is not backportable to previous releases.
Note 356583.1 has been linked to this scenario.
Bug 3517013 (Unpublished)
Abstract :OPEN DB RESETLOG AFTER FLASHBACK DB FAILS ORA-600 [KCLCHKBLK_4], [1904]
Symptoms :
1) When restored the database from backup and did an incomplete recovery.
2) Opened the database with resetlogs.
3) After opening the database, you start getting following errors:
ORA-00600 [kclchkblk_4]
ORA-00600 [2662]
4) Stack trace is:- kclchkblk kcbzib kcbgcur ktfbhget ktftfcload
Cause :
1)
Error, ORA-600[KCLCHKBLK_4], is signaled because the SCN in a tempfile block
is too high. The same reason caused the ORA-600[2662]s in the alert logs.
2)
This issue is because the tempfiles may not get reinitialized during open
resetlogs.
Patch details :
Currently there is no one-off patch available for any versions/platforms.
Note 275902.1 has been linked to this scenario and solution
given under this note.

Many other bugs were filed with development for this issue.
Those bugs are not progressed due to
— Lack of response from the customers
— one-time occurances
— Vendor OS Problem