Re: RAC in NAS

strace the dbwriter process and look for io_submit() vs. pwrite().

disk_async_io applies to RMAN backups.

I hope it's not answered further down, i am still reading this.

--
Christo Kutrovsky
Senior Database/System Administrator
The Pythian Group - www.pythian.com
I blog at http://www.pythian.com/blogs/


On 7/28/06, Mark Brinsmead <pythianbrinsmead@xxxxxxxxx> wrote: 
Nuno,

   You are only *half* right, I fear.

   Actually, at least one of -- and perhaps both -- MetaLink and Werner
describe more or less
the same caveat, that is:  applications *other* than Oracle may be
responsible for the non-zero
counts in slabinfo.

   But here's the half where you're wrong.  (Or at least, not completely
"right".)   Those hypothetical
"other" applications can just as easily be resposibly for the *increases* in
the slabinfo stats as
they are they would be for "non-zero" values.  After all, those values had
to increase from zero
*some* *time*, didn't they.  ;-)

   Anyway, you are correct about this:  I am still unaware of a simple way
to prove *conclusively*
that a database is actually using Asynch I/O.  But (on a good day) I now at
least know how to
prove that it is *not*.  (Or at least that *no* databases *are*.  Sadly,
multiple database on the
same host muddy the water even more.)

   If anybody out there can tell me of a simple (and reliable) test that
*proves* a database is using
Asynch I/O, I'd like to hear about it...

   In the meantime, it has (happily) met my purposes to be able to prove the
negative.



On 7/28/06, Nuno Souto <dbvision@xxxxxxxxxxxx> wrote:
> Mark Brinsmead wrote,on my timestamp of 28/07/2006 1:39 PM:
>
> > (Almost) just for chuckles, I opened an SR with Oracle support, asking
> > questions like "how can
> > I test whether my DB is doing Asynch I/O on Linux?" and "knowing that
> > Asynch I/O is unsupported,
> > what are the risks of doing so anyway?".  After almost two weeks, the
> > questions are unanswered,
> > even though I was able to answer them myself with less than an hour of
> > surfing Metalink and Google.
>
> yes, there is a note in metaclick explaining how to check.
> But it's not complete, neither is werner's site:
> you check for those counters in /proc/slabinfo being
> non-zero *AND* changing in value when you startup Oracle!
> There might be *other* software around already using aio
> and just having them as non-zero is not enough to say
> Oracle is using it.
>
> DAMHIKT...
>
>
> --
> Cheers
> Nuno Souto
> in (finally) sunny Sydney, Australia
> dbvision@xxxxxxxxxxxx
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>



--
Cheers,
-- Mark Brinsmead
   Staff DBA,
   The Pythian Group
   http://www.pythian.com/blogs



--
Christo Kutrovsky
Senior Database/System Administrator
The Pythian Group - www.pythian.com
I blog at http://www.pythian.com/blogs/
--
http://www.freelists.org/webpage/oracle-l


Other related posts: