Re: Got the darn buffer busy waits under control, at last!

Nuno-

If latch free was not a significant wait event in the first place, then the spin count exercise is a wasted effort. You are not addressing the real problem, which is the discovery and relief of the worst bottleneck.

I would suggest you remove the spin count parameter altogether so as not to cloud issues in the future. Yes, you achieved an apparent change in latch behaviour. On the other hand, you cannot really say you improved the instance's performance in any measurable way. Later, when you tune the real issues, the dynamics of the instance may change such that the spin count thing bites you in the ass.

• ricky

Nuno Souto wrote:
>
> In article <3D094FC2.82A095CE_at_more.net>, you said (and I quote):
> >
> > I would be curious to see how the instance was affected otherwise. Was
> > latch_free a major wait event prior to the parameter adjustment? Has
>
> Nope.
>
> > overall latch_free wait time actually gone down as a result? What about
>
> Yes.
>
> > sleeps for other latches? Did we trade cache buffers chains latch
> > performance for library cache latch?
> >
>
> Nothing much.
>
> > It would be useful to see before and after statspack reports rather than
> > have to rely on anecdotal evidence. Nuno- you have statspack running on
> > that thing? Post a zipped 30 minute busy period snap if you have one.
> >
>
> Don't have statspack, only the older bstat stuff.
> I've got a few from before and am collecting afters.
> It's EOM for them now, had to wait a coupla days until things settled
> back to normal load. Otherwise it's apples and oranges.
> Will post both this weekend.
>
> --
> Cheers
> Nuno Souto
> nsouto_at_optushome.com.au.nospam
Received on Sun Jun 16 2002 - 20:13:08 CDT