Clear.
Sure, the cache is nice. EMC are the gods of cacheEventually, the writes
have to be made permanent. What happens when the cache is full? Right, EMC
blocks io until most of the io has gone to disk. This leaves the poor user
waiting, and waiting, and grow a beard.
Symmetrix is limited
Why worry about how it gets written. The cache will make any implementation sing like
a bird. And if you think
you are having problems just add more cache and that will take care of it.
- As told to us by a past EMC Sales Weenie
-Original Message-
Sent: Friday, October 03, 2003 6:50 AM
To
This could be true on the request level. But it will not stop it from
pipelining other requests on the spindle level, so that when spindle A is
done for request 1 it works on request 2. This increases the overall
throughput of the system.
Then using Async IO should help here.
I'm talking about EMC
Sorry Matt,
Just checked with EMC netherlands, unfortunately, I'm right.
About the cache: a shared cache is easily flooded with large io's.
Reply-To: [EMAIL PROTECTED]
To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
Date: Fri, 03 Oct 2003 06:01:00 -0800
Hi Hans,
Absolutely not true,
Hi Hans,
Absolutely not true, and has not been true for a long time. Writes to an
EMC never go directly to disk anyway, and when they do go to disk is purely
determined by the microcode algorithms, and will often have nothing to do
with stripe layout at all.
Thanks,
Matt
- Original Message
Hi All,
Today I saw an archived thread on orafaq about striped volumes in an EMC
Symmetrix. Gaja mentioned that writing to a striped volume is performed in a
sequential fashion i.e. spindle B will not start writing block 2 before
spindle A has completed writing block 1.
Is this still true for