Sent: Thursday, August 21, 2003 1:49 AM
>
>
> > so in 9i dbms_rls increases the soft parses?
> > >
> > > From: Connor McDonald <[EMAIL PROTECTED]>
> > > Date: 2003/08/20 Wed AM 11:21:59 EDT
> > > To: Multiple recipients of list ORACLE-L <[EMAIL PROTECT
gt; >
> > From: Connor McDonald <[EMAIL PROTECTED]>
> > Date: 2003/08/20 Wed AM 11:21:59 EDT
> > To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
> > Subject: Re: Re: Row level security and latch waits
> >
> > My understanding was that
> <[EMAIL PROTECTED]To: Multiple recipients of
> list ORACLE-L <[EMAIL PROTECTED]>
> uk> cc:
>
so in 9i dbms_rls increases the soft parses?
>
> From: Connor McDonald <[EMAIL PROTECTED]>
> Date: 2003/08/20 Wed AM 11:21:59 EDT
> To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
> Subject: Re: Re: Row level security and latch waits
>
> My underst
My understanding was that the rls predicate was added
at parse time (hence the importance of the contexts
and avoiding things like 'sysdate')
But also if I remember correctly, this behaviour was
changed in v9 to process the security function with
each execution (and hence probably increase the am
- Original Message -
> - Use context values within predicates, as bind variables NOTE THIS
> LINE.>
None of that implies it's not appending a where clause like it
was explained.
It uses the context values (if that's the case) in a function
which then returns a string. Tha
Hi Raj,
Vadim is right and if RLS is implemented without application context then
there would be hard parses especially if there are literals in the policy
function . We have a client warehouse with about 500 users and 300 tables
and FGAC has been working very well and we do not see latch waits.
T
To: Multiple recipients
of list ORACLE-L <[EMAIL PROTECTED]>
Sent by: cc:
[EMAIL PROTECTED]Subject: Re: Re: Row level
security and latch waits
.c
>
> >To: Multiple recipients of list
> ORACLE-L <[EMAIL PROTECTED]>
> Sent
[EMAIL PROTECTED]Sub
its just appending a where clause. its not binding it.
im not familiar with contexts. never worked with them. someone correct me if im wrong
here? Could have sworn i read that somewhere.
i looked up application contexts. they appear to be handled differently.
am i wrong?
>
> From: [EMAIL PROT
im fairly certain its because DBMS_RLS doesnt use bind variables now. latch waits
indicates plus your 100 concurrent users.
you can check this by going to v$sqlarea and checking for similiar sql_text
statements. if they are there, then your not using bind variables. there is a query in
tom kyte
12 matches
Mail list logo
