RE: Unexplained table growth

2002-01-18 Thread Jesse, Rich
Hmmm...the AVG_ROW_LEN is the same for both: 72. The part that sticks out is the production table has a NUM_FREELIST_BLOCKS of 28569, while test is at 5. Any ideas as to how this could be??? Almost like the TRUNCATE didn't clear out the freelist??? Or am I missing/misinterpreting something her

Re: Unexplained table growth

2002-01-18 Thread Stephane Faroult
If I were you I'd try to analyze both tables (prod and test), posibly on a small sample, to check whether the average row length is identical. The 'comment' field, which can be close enough to 60%, may be NULL in your test database and full in most cases in prod. Something else one could think of

RE: Unexplained table growth

2002-01-18 Thread Vadim Gorbounov
Hi, Jesse, Didn't your dev use parallel insert on production to speedup things? This may cause different segment allocation HTH Vadim -Original Message- Sent: Friday, January 18, 2002 12:41 PM To: Multiple recipients of list ORACLE-L Hi all, 8.1.7.2.0 on HP/UX 11.0 -- A developer tes

Unexplained table growth

2002-01-18 Thread Jesse, Rich
Hi all, 8.1.7.2.0 on HP/UX 11.0 -- A developer tested a table recreate in our test DB. Here's the scenario: CREATE TABLE MY_BIG_TABLE ( FISCAL_YEAR NUMBER (5), PERIODCHAR (2), ACCOUNTNO CHAR (12), TRANSTYPE CHAR (2), TRANSQTY FLOAT, TRANSAMNT FLOAT,