Instance terminated by PMON

[pr_srinivas]

DB Version  : 11.2.0.2.0
DB Archive Log Mode : No - Archive Log Mode
DB Backups : Not available
Yesterday, There was a STORAGE ABNORMAL SHUTDOWN HAPPENED. after STORAGE WAS restored.
The Oracle Database is crashing down in few minutes as soon as we start the Database.

[Former Member · Accepted solution]

Hi,

In this case we can do one thing. creating new undo tablespace as below.

create the tablespace with this SQL

create undo tablespace undotbs2 datafile ‘/u02/oracle/oradata/test/undotbs2.dbf’ size 500m;

Alter the database to use the new UNDO tablespace.

alter system set undo_tablespace=undotbs2 scope=both;

Then take the original UNDO tablespace offline

alter tablespace undotbs1 offline;

Next drop the corrupted segment.

drop rollback segment <corrupted segment>;

Then finally drop the old UNDO tablespace

alter tablespace undotbs1 offline;

Regards,

Venkata S Pagolu.

[pr_srinivas · Accepted solution]

Tons of Thanks to  You - Venkat (Venkata Satyanarayana Pagolu)

We used your advise & recovered the system.
Rgds
PR

Oracle Note :
How to Change the Existing Undo Tablespace to a New Undo Tablespace [ID 431652.1]

Metalink Login Credential SAP Note :
Note 758563 - Oracle Metalink access for SAP customers

[stefan_koehler]

Hello Srinivas,

that looks pretty bad.

Block recovery completed at rba 2703.136.16, scn 4.1473908910 
Errors in file /oracle/Q11/saptrace/diag/rdbms/q11/Q11/trace/Q11_smon_975068.trc: 
ORA-01595: error freeing extent (3) of rollback segment (3)) 
ORA-00607: Internal error occurred while making a change to a data block 
ORA-00600: internal error code, arguments: [kdBlkCheckError], [2], [1017225], [38504], [], [], [], [], [], [], [], []
Corrupt Block Found 
         TSN = 1, TSNAME = PSAPUNDO 
         RFN = 2, BLK = 1017225, RDBA = 9405833 
         OBJN = 0, OBJD = -1, OBJECT = , SUBOBJECT =  
         SEGMENT OWNER = , SEGMENT TYPE = 

It seems like your undo tablespace is corrupted. You need a backup to get it working again on your own.

Maybe Oracle itself is able to fix that corruption, but you need to open a SR for that.

Regards

Stefan

[pr_srinivas]

Alert Log File as below

alert_Q11.txt.zip