>
> I've always thought of "sweep" as the servers rather rude way to

clean up

> when you force dirty applications on it for some time or you fail to

do some

> standard maintenance on the database in a routine, timely fashion.
>
> I disable sweep (it maybe a left over from vesion 5.5 when sweep was

broken)

> but i check the server stats from time to time. If the application

is clean

> and the computers running it are stable, then sweep may never be called
> anyway, even if it's set at 5,000.

We have also disabled the sweep function.
Also we take a look at the DB stats every once in a while.
The transactions are always close top each other (at least when all
users log out at night time).
We have made a small rutine to sweep the DB every night at 01.00.
This normally takes 2 minutes.
But once in a while the users have aborted a rather big update
(rollback) which has made the gab between transactions rather big.
And here the sweep takes over.

>
> I do a backup and restore about every six months. This is after what

I call

> good, healthy, moderately heavy user traffic by up to 15-20 clients

8 hours

> a day 5 days a week on a database of about 250Mb. I might even

overlook a

> restore til 8 or 9 months.

We actually never do a backup restore unless something crashes.

>
>
> > > >Could the size of the table (~1M recs) significantly affect the
> > speed of
> > > >returning records read by index?
> > >
> > > It could. How often do they run SET STATS to rebalance the indexes?
> >
> > Is this something that should be done every once in a while ?
> > What does it do
> >
>
> I have an adminstrative button in my app which rebuilds the indexes.

I can

> set it off when I like. I have never really noticed a difference in app
> performance after such a rebuild.
> If you have the correct indexes already, it will perform well. I was

running

> one application for about 2 years. We did an upgrade - both hardware and
> software and I rebuilt the database (not completely). After

installing all

> the new stuff, It was noticeably faster - everyone was pleased. Then

one day

> about 3 months later, I was poking around and noticed that an index

I was

> used to seeing wasn't there (my fault). One index on one table. I put it
> in.. and boy did it make a difference. I was asked by the users what had
> happened. they all noticed. They never notice when I rebuild the

indexes.


This we also experienced.
We have a lot of indexes in oour DB.
It works well, and noone ever notices anything.

Except the first time, when I created some more indexes, they all also
said it went faster.

Thank you

Michael

>
>
> Alan