home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.databases.oracle
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!decwrl!elroy.jpl.nasa.gov!sdd.hp.com!zaphod.mps.ohio-state.edu!rpi!psinntp!psinntp!gdstech!un
- From: un@gdstech.grumman.com (Un Fu)
- Subject: Thanks for the solution : ) to "extent allocation?" & "Help! Oracle problem or my problem?"!
- Message-ID: <1992Nov2.191506.23604@gdstech.grumman.com>
- Sender: un@gdstech.grumman.com (Un Fu)
- Organization: Grumman Data Systems-Woodbury
- Date: Mon, 2 Nov 1992 19:15:06 GMT
- Lines: 32
-
- I would like to thanks all the responsers to my posting of
- the following two qestions on the net:
-
- "extent allocations"?
- "Help! Oracle problem or my problem?"
-
- Andy Finkenstadt, Scott Stephens => Thanks for the dummy table solution!
- I am so glad there is a way to get around
- the problem. I would never figure it out
- myself.
-
- To the responsers that told me there is no way out, here is the work
- around:
-
- # of bytes
- |
- CREATE TABLE BIG_TEMP (PARAM1 CHAR(10)) STORAGE (INITIAL 1000000 MINEXTENTS 1);
- DROP TABLE BIG_TEMP;
-
- As Scott and Andy stated to me, Oracle first looks for exact fit, then best
- fit, and then coalesce of smaller, CONTIGUOUS, extents when it allocates
- extents for object creation. The DUMMY table creation above will force
- the coalesce of smaller, contiguous, extents into one large extent.
- I tried the solution and it worked, awesome! You might have to coalesce
- the small extents in several passes if you have many. I ran into an
- insufficient row_cache_enqueues error when I tried to coalesce 5000 blocks
- at once.
- --
- ****************************************************************************
- Un Un Fu Internet: un@gdstech.grumman.com
- Disclaimer: I am solely responsible for what I am saying here!
- ****************************************************************************
-