Now, I'm curious again...time to do some investigation.
From initial tests, from:
BLOCK_SIZE: 16384
DB_CACHE_SIZE: 12582912
DB_FILE_MULTIBLOCK_READ_COUNT: 16
Created a table called t1 (object id = 29230) (190 Blocks);
Created a table called t2 (object id = 29231) (190 Blocks);
Startup
Shutdown
SELECT COUNT(*) FROM t1;
SELECT COUNT(*) FROM v$bh where obj = '29230';
COUNT(*)
---------
17
Which is db_file_mulitblock_readcount + 1 (extra one is a buffer header)
SELECT COUNT(*) FROM t2;
SELECT COUNT(*) FROM v$bh where obj = '29231';
COUNT(*)
---------
17
Then I went back to see if the other T1 fts data was present...
SELECT COUNT(*) FROM v$bh where obj = '29230';
COUNT(*)
---------
1
Only one block in the BufferCache was present.
Did
SELECT block#, objd from v$bh
order by block#;
And the blocks toward the LRU end contain the objects' blocks.
Last edited by grjohnson; 03-13-2003 at 02:25 AM.
OCP 8i, 9i DBA
Brisbane Australia