on 04-02-2008 2:53 PM
I have a materialised view in an external database that gets refreshed via a materialised view log on AUSP.
Database is Oracle 9.2.0.7 on TRU64.
R/3 version is 4.7 ext110
I've been looking at the number of records in this log (MLOG$_AUSP), total fluctuates between 0 and 1200 records.
Not much!
However, if I try and do a select count(*) from sapr3.MLOG$_AUSP it takes over a minute.
The execution plan is:
COUNT(*)
----------
0
Elapsed: 00:01:42.60
Execution Plan
----------------------------------------------------------
0 SELECT STATEMENT Optimizer=CHOOSE (Cost=86389 Card=1)
1 0 SORT (AGGREGATE)
2 1 TABLE ACCESS (FULL) OF 'MLOG$_AUSP' (Cost=86389 Card=506
)
Statistics
----------------------------------------------------------
0 recursive calls
0 db block gets
569248 consistent gets
529854 physical reads
0 redo size
517 bytes sent via SQL*Net to client
652 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed
I don't get it. Why could there be so many reads if the record count is so low?
This is also affecting my database buffer quality.
Any help would be appreciated.
Well, it would appear a simple reorg has fixed the problem!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
86 | |
10 | |
10 | |
9 | |
6 | |
6 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.