Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Locking: Matching up sessions

Re: Locking: Matching up sessions

From: Stephane faroult <sfaroult_at_oriolecorp.com>
Date: Fri, 03 Sep 1999 00:49:56 -0700
Message-ID: <37CF7DA4.24C3@oriolecorp.com>


Ben Ryan wrote:
> Imagine three sessions are each updating a single row in
> the table and none of them have yet commited the change.
> Call the sessions A, B and C. Let us say
> A is updating r1
> B is updating r2
> C is updating r3.
>
> Imagine two more sessions called x and y. Which,
> are both trying to a update a single row in the table
> and they were both initiated after sessions A, B and C
> had acquired their locks. Lets us say that
> x is trying to update r1 and
> y is trying to update r2.
>
> By querying v$lock.id1 I can see how to say that
> some combination of sessions A, B and C are blocking
> sessions x and y.
>
> Is there anyway to discover that
> A is blocking x and
> B is blocking y

If I remember well, in V$LOCK you have not only locks which are held but also those which are requested. There is in $ORACLE_HOME/rdbms/admin (I assume a Unix box!) a script or two which give a better depiction of locking than the raw V$LOCK. I think there are utlsomething.sql scripts, assuming Unix once again try a grep -i blocking *.sql on the said directory I am sure there is something. You can also search for scripts on a number of sites; we have a monlocks.sql for lock monitoring on our site although to be honest I don't remember what it displays exactly. --
Regards,

  Stéphane Faroult
  Oriole Corporation
  Performance Tools & Free Scripts



http://www.oriolecorp.com, designed by Oracle DBAs for Oracle DBAs
Received on Fri Sep 03 1999 - 02:49:56 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US