Re: Locktype and performance
From: Apostolos Dimopoulos (adimop_at_dreamtech.gr)
Date: 12/05/03
- Previous message: Apostolos Dimopoulos: "Re: Locktype and performance"
- In reply to: George Christoforakis: "Re: Locktype and performance"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Fri, 5 Dec 2003 12:38:54 +0200
Hi George,
Thanks for the info. Still the question remains why when I am the sole user
of the database and running a fairly large view I get a performance of about
0.2 secs using adUseServer / adOpenForwardOnly / adLockOptimistic and about
2 secs using adUseServer / adOpenForwardOnly / adLockReadonly...
Apostolos
"George Christoforakis" <nospam@nospam.info> wrote in message
news:3fcb082f$1@newsgroups.borland.com...
> Hi Apostolos,
> OptimisticLock is designed like that. It doesn't lock at reading.
> ReadonlyLock, puts record it in a mode that it can not be updated by
> anyone. So eventually I assume that it somehow locks it. Whenever there is
a
> lock situation, the speed gets slower.
> Also make sure that these settings do not change during runtime. If ado
> determines that it can not use (eg clUseServer) for any reason, then it
will
> change it automatically. The same applies on locks.
>
> George Christoforakis.
>
>
> "Apostolos Dimopoulos" <adimop@dreamtech.gr> wrote in message
> news:3fc7632e@newsgroups.borland.com...
> > Hi,
> >
> > I am using the ADO controls with Delphi 7 and SQL Server 2000 to
populate
> a
> > control with data from a large table. Even though I would expect the
> > combination of adUseServer / adOpenForwardOnly / adLockReadonly to
offers
> > the best performance, I see that the best performance - with great
> > difference is obtained when using adUseServer / adOpenForwardOnly /
> > adLockOptimistic or adLockBatchOptimistic.
> > Also adLockOptimistic takes no locks for reading as expected but
> > adLockReadonly takes select locks.
> > Any explanation regarding this performance behaviour?
> >
> > Thanks in advance,
> > Apostolos
> >
> >
> >
>
>
- Previous message: Apostolos Dimopoulos: "Re: Locktype and performance"
- In reply to: George Christoforakis: "Re: Locktype and performance"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|