[Oracle] [corruption] How to investigate when ORA00600[KDSGRP1] occurs

Essentially, this is probably caused by a bad block, so you need to investigate the bad block condition in the associated table:

Excerpt of trace file
============================
2017-08-18 09:23:04.323
Dbkeddefdump (): Starting incident Default dumps (flags = 0x2, level = 3, mask = 0x0)
[TOC00009]
-----current SQL Statement to this session (sql_id = 7SNQRQ5TH8DDH)-----
With tmp001 as (
...
(short)
...

Select
COUNT (*)
From
TMP001 2
where 1 = 1

and ROWNUM <=: 9
============================


Please check the table above for damage respectively:

1. Check whether there is a problem with the structure for each table:

Example:

Analyze table table name validate structure;
Analyze table table name validate structure cascade;

2. Check the data files to which the above tables belong:

Example:

---How to check the name of the data file belonging to the table:

Sql> select file_name from Dba_tables T1, Dba_data_files T2
Where T1.tablespace_name = T2.tablespace_name
and T1.owner = ' owner name ' and t1.table_name = ' table name ';

---How to check for corruption in data files:

$ Rman Target/
rman> backup Check logical validate DataFile ' the acquired data file name ';

After checking, please check whether there are damage in V $ database_block_corruption view:

Sql> SELECT * from V $ database_block_corruption;


If the error occurs and the breakage can be confirmed by the above check, it separates whether it's temporary damage on M Emory
Restart the database, please check again.

[Oracle] [corruption] How to investigate when ORA00600[KDSGRP1] occurs