Topic: Performance & Tuning >> UNDO Tablespace gets corrupted time and again
|
Disclaimer: The purpose of all dbapool forums including OCP and Other Oracle Certification forums is to help each other with specific issues but not to share dump and copyrighted exam content, materials or intellectual property.
You may review the entire Oracle Certification Program Candidate Agreement online Here.
|
|
|
|
| Title: UNDO Tablespace gets corrupted time and again |
| PRANABH |
Posted: Oct 16, 2006 01:30:37 AM |
Total Post: 4
Joined: Sep, 2006
|
I am finding reasons why in the database i am handling UNDO tablespace gets corrupted time and again so that i can rectify the
problem.
For this
1.I have shanged undo_retention parameter to 40 so
that there is not much undo retention required which
may have caused to break down of undo segments.Earlier
it was 10800.
2.I have added a datafile to undo tablespace with
autoextend on and checked parameter undo_management.It
is AUTo so that the error may not be becuse of less
undo space or expansion of undo previous datafile.
3.I have decreased fast_start_mttr_target from 300 to
50 so that checkpoint is fired after approx. so much
time and recovery time is reduced,causing less undo
utalization.
4.I am looking for any network errors which may have
caused distributed processing haults during updation
causing undo segment failures.I saw some warning
messages in windows server regarding network error.
Though if a transaction is breaked in between updation
without commit or rollback the Oracle background
processes should bring the same in consistent state
but I think that due to large updations there is some
problem with undo segments.This is under observation
and now i will be updating it under my supervision
and not by batch processing so that i can
know wether it is the root cause of problem.
Pls give your suggesions on above points.
Thanks and Regards
Pranabh
|
|
|
 PP |
| Posted: Oct 26, 2006 01:28:42 AM |  |
|
Total Post: 11
Joined: Oct, 2006
|
1) Depends on your requirment of flashback support.Although,this may not be related to any undo corruption you are facing.
2) Adding a datafile with autoextend on is dangerous. Instead of becoming a solution to your present problem,it may become a potential cause for furhter aggrevating it.If there is a short of space in both the filesystem and the undo tablespace,an automatic extend of the datafile may cause an os error which, in turn, can cause "corruption" in the tablespace ( I have faced this).Instead, if the auto extend is turned off and the tablespace is full, it will give only an oracle error and certainly not corrupt any tablespace( I can say this with some confidence for Oracle 9i). My advise to you will be to check whether other datafiles of this tablespace has auto extend on or not and turn them off and watch whether the corruption repeats itself or not.
3) fast_start_mttr_target is related to redo and not undo. Undo totally depends on whether transactions are being commited or not, and undo retention. Decreasing fast_start_mttr_target will affect performance of the database.
4)Cannot comment on this without knowing your environment.What errors are you getting btw? Can you post the error number and descriptions?
|
|
|
|
|
| Time Zone: EDT |
Send this thread to your friend |
|
|
|
|
Forum Rules & Description
Who Can Read The Forum? Any registered user or guest
Who Can Post New Topics? Any registered user
Who Can Post Replies? Any registered user
|
| |
Get FREE Magazines
|
Top 10 Forum User
|
| Murtuja Khokhar | 814 |
| Mohammed Taj | 694 |
| Jayanta Sur | 479 |
| Vigyan Kaushik | 386 |
| positive fanatic | 361 |
| Gitesh Trivedi | 322 |
| Gopu Gopi | 242 |
| neeraj sharma | 228 |
| Ramesh Jois | 226 |
| snehalatha p | 169 |
|
|
