Hi Pablo,
The TM and TX waits are probably normal application tuning issues. You'll find
some tips about reducing ST enqueue waits on the Ixora web site, MetaLink and
elsewhere. The CU and SQ waits are relatively few and probably relatively brief,
nevertheless there can be performance issues with
Thanks for the answer Steve.
--- Steve Adams <[EMAIL PROTECTED]> escribió: >
Hi Pablo,
>
> The TM and TX waits are probably normal application
> tuning issues. You'll find
> some tips about reducing ST enqueue waits on the
> Ixora web site, MetaLink and
> elsewhere. The CU and SQ waits are rel
see p2text and p3text for more.
Sounds like a one-time long-held table/row lock to me.
If you can, bounce the instance and recheck for reoccurence
hth
Ross "I wish I could do statistics" Mohan
-Original Message-
Sent: Tuesday, May 15, 2001 2:50 PM
To: Multiple recipients of list
Hi Diego,
Without going into details, an 'enqueue' wait is mostly due to a
user/program initated transaction lock and I see it a lot in Financial
databases (I see you are on 10.7?). I deduce you were looking at
V$SYSTEM_EVENT - If a user locked one row (probably queried a row in Forms,
inadvertl
Thanks for answering Unal, John.
John, you were right about the query. I've corrected it. And I'm not getting
"T[" and "CK" anymore,now I get TX and CI.
I've also done further investigation and I also know now what p2 and p3
mean.
This is the updated data: (from the dumps)
--
Hi Diego,
No, those are block buffers. When a segment is dropped, truncated or shrunk
(normally a rollback segment) then a "reuse block range" cross instance call is
needed to flush the unwanted blocks from cache. Similarly, before a parallel
direct read a "checkpoint block range" or "checkpoint
Hi Diego,
If lock type is 'TX', rollback segment number = trunc(p2/65536)
and slot number = p2 - 65536*trunc(p2/65536).
May I know which table you are looking at to get the following
detailed info about enqueue waits?
Enqueue Stats
--
TY GETS WAITS
s
> @ http://www.ixora.com.au/
> @ http://www.christianity.net.au/
>
>
> -----Original Message-
> From: Diego Cutrone [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, 17 May 2001 4:11
> To: Multiple recipients of list ORACLE-L
> Subject: RE: enqueue waits -- CI
>
sibility of parallel server, and the DBWn processes in
all
> instance need to flush the cache in their own instances.
>
> @ Regards,
> @ Steve Adams
> @ http://www.ixora.com.au/
> @ http://www.christianity.net.au/
>
>
> -----Original Message-
> From: Diego Cut
Jay,
You may want to refer to Metalink Doc Id #29787.1 and 34566.1. Those will explain
what p2, p3 are in an enqueue wait. Unfortunately, those will not discuss p2, p3 for
UL :(
I know nothing about Oracle Apps (11i). May be John K. could comment on that.
- Kirti
-Original Message-
Jay,
Does this come from the alert manager? Do you have any of those new-fangled
11i modules (or should I call the 'mangled'!!). You could use the script
below (adapted from Govind who posted this a few days back)
set pages 100
column sid_serial format a10 heading "Sid/Ser#"
column username forma
Here is the output from the query:
Sid/Ser# DB/OSUser StartTime Mins Rbs NameBlks Status
-- --- -- -- -
14/106 APPS/applmgr12/06/02 01:29:32 518 RBS2 1 ACTIVE
33/537 APPS/applmgr12/06
OTECTED]]
> Sent: Friday, December 06, 2002 8:51 AM
> To: Multiple recipients of list ORACLE-L
> Subject: RE: Enqueue Waits in Oracle Financials
>
>
> Here is the output from the query:
>
> Sid/Ser# DB/OSUser StartTime
John,
Yes, the PMON method is lock.
The ICM runs 1 process with a sleep time of 30 seconds and a null value for cache
size.All of our managers sleep at least 30 seconds, with the exception of a
Service Manager, which is null.
I'll probably end up logging a TAR.
Thanks,
Jay
>>> [EMA
John,
I know this is an old topic, but Oracle Support proposed a patch and we finally got
it tested. It looks like this enqueue wait goes away with the application of
concurrent processing rollup patchset C 2385942.
Thanks,
Jay Hostetter
Oracle DBA
D. & E. Communications
Ephrata, PA USA
Thanks for the TX information Jeffery.
But my question about CI enqueue still remains.
Here's the query to get enqueue waits statistics, it's from Steve Adams'
site.
select
q.ksqsttyp type,
q.ksqstget gets,
q.ksqstwat waits
from
sys.x_$ksqst q
where
q.ksqstget > 0
/
thanks again
CI - Cross-instance Call Invocatiom
UL - User-defined locks
[EMAIL PROTECTED] wrote:
>
> The following is from a Statspack report from an 8.0.5 database on a Sun
> server during a load test.
> Can anyone explain what type of enqueues these are and where I can find some
> doco on them? I did loo
17 matches
Mail list logo