Re: What to keep in ASM?

Well, read IO hotspots based on single (or few) allocation units (1
MB) are unlikely as database cache will shield IO subsystems from
that. Write IO hotspots can be avoided with less aggressive
checkpointing.

Indeed, there is always a chance that extents alternate so that most
active data goes to one disk but this is not very likely. Is it?


On 2/26/07, K Gopalakrishnan <kaygopal@xxxxxxxxx> wrote:

Don,

It is slightly incorrect. ASM does not work at block level. It works
at extent level or allocation unit level (1MB most cases) and
hotblocks exist in ASM too. You may have to tweak the block level
parameters to avoid hotblocks.. You may want to use X$KSLHOT to nail
down them.


On 2/26/07, Don Seiler <don@xxxxxxxxx> wrote:
> My understanding is that ASM will also shift data around to avoid the
> "hot block" scenario, avoiding I/O contention.  It wasn't explicitly
> stated in the passage I pasted, hopefully my understanding is correct.
>
> Don.
>
> On 2/26/07, Alexander Fatkulin <afatkulin@xxxxxxxxx> wrote:
> > > Am I giving this feature too much credit?
> >
> > ASM spread extents in proportion to the disk size regardless of the
> > speed and workload characteristics of the underlying hardware (at
> > least I never saw the opposite).
> >


--
Best Regards,
K Gopalakrishnan
Co-Author: Oracle Wait Interface, Oracle Press 2004
http://www.amazon.com/exec/obidos/tg/detail/-/007222729X/

Author: Oracle Database 10g RAC Handbook, Oracle Press 2006
http://www.amazon.com/gp/product/007146509X/
--
http://www.freelists.org/webpage/oracle-l






--
Best regards,
Alex Gorbachev

http://www.oracloid.com
--
http://www.freelists.org/webpage/oracle-l

Other related posts: