Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: Truncate command internal behavior
Thanks for the responses.
"See bug 5177241 for some more info. Looks
like Oracle is dancing around issues with CKPT/DBWR communication in
10g, the RO enqueue is used to syncronize that, which is probably where
you are seeing fast object reuse waits"
We are using 10.2.0.2 and this bug was in 10.1 and suppose to be fixed in 10.2
I will open a Tar on this issue and let see what Oracle have to say about this truncate behavior.
Thanks
--Harvinder
On Oct 17, 9:59 pm, wagen..._at_yahoo.com wrote:
> In one of our db's, 9.2.0.4 on solaris 9 db_cache-size 4GB, when we did
> a truncate of relative small table < 50M, the truncate command would
> hang with the rdbms ipc message (wait event). I gather that this was
> due to DBWR doing async writes of the dirty buffers to disk, but was
> slow (io subsystem problem ?).
>
> we moved to 9206 and this behaviour disappeared.
>
> could you do a alter system checkpoint before the truncate command and
> see if this makes any difference.
Received on Wed Oct 18 2006 - 08:57:49 CDT