Re: What is Parallel DML?

Read Jonathan Lewis' post on the subject, and you will see that my argument is not nonsensical, nor wrong-headed, nor circular.

Arrogant shits who refuse to look facts in the face annoy me.

You annoy me.

You can complete the syllogism.

HJR "Billy Verreynne" <vslabs_at_onwe.co.za> wrote in message news:9jj16k$dpk$1_at_ctb-nnrp1.saix.net...
> Howard J. Rogers wrote
>
> >That doesn't get around the fact that at any one moment, only a single
> >process can be serviced by a single CPU, regardless of capacity.
>
> What moment? Per second? Many processes can be run on the CPU per second.
 Per
> milisecond? Per nanosecond? This argument is nonsensical.
>
> A single CPU can service many processes simultanously. And if you look
 closely
> at the design of the new generation CPUs, even the "fact" that a CPU can
 only
> executed a single instruction in a single point in time, no longer holds
 true.
>
> > >effective use, but only one of those slave process can actually have
 time
> > >on the CPU at any given moment. And *that's* the reason my query
 slowed down
> > >with a degree of parallelism of 2 -having to perform a context switch
> > >between two processes on the one CPU is expensive.
>
> Context switching is _not_ the issue.
>
> > >Threads are a different game again, since there is no context switch
 when
> > >switching between threads within the one process.
>
> So PQs are not threads on a platform like NT? Or because PQ are threads on
 NT,
> it works on a single CPU NT machine, but not when running Linux with
 Oracle on
> that exact same machine? You are arguing in circles.
>
> The fact is that PQ does work on a single CPU platform. Period. From a
 200Mhz
> Pentium running Oracle 7.3 to a massive parallel cluster of single CPU
 Unix
> nodes running Oracle Parallel Server. Been there. Done that. Have the
 t-shirt.
>
> >Your mileage may vary: mine never has.
>
> Yeah, that is always a problem with closed minds that never want to
 understand
> and tweak for better mileage.
>
>
> --
> Billy
>
>
>
>
Received on Tue Jul 24 2001 - 04:19:38 CDT