At 04:15 PM 12/27/2005, Michael Stone wrote:
>I don't understand why you keep using the pejorative term "performance
>hit". Try describing the "performance characteristics" instead.
pe·jor·a·tive ( P ) Pronunciation Key (p-jôr-tv, -jr-, pj-rtv, pj-)
Tending to make or become worse.
RAID 5 write performance is significantly enough
less than RAID 5 read performance as to be a
matter of professional note and concern. That's
not "disparaging or belittling" nor is it
"tending to make or become worse". It's
measurable fact that has an adverse impact on
capacity planning, budgeting, HW deployment, etc.
If you consider calling a provable decrease in
performance while doing a certain task that has
such effects "a hit" or "bad" pejorative, you are
using a definition for the word that is different than the standard one.
>Also, claims about performance claims based on experience are fairly useless.
>Either you have data to provide (in which case claiming vast experience
>is unnecessary) or you don't.
My experience _is_ the data provided. Isn't it
convenient for you that I don't have the records
for every job I've done in 20 years, nor do I
necessarily have the right to release some
specifics for some of what I do have. I've said
what I can as a service to the
community. Including to you. Your reaction
implies that I and others with perhaps equally or
more valuable experience to share shouldn't bother.
"One of the major differences between Man and
Beast is that Man learns from others experience."
It's also impressive that you evidently seem to
be implying that you do such records for your own
job experience _and_ that you have the legal
right to publish them. In which case, please
feel free to impress me further by doing so.
>>Said RAID 10 array will also be more robust
>>than a RAID 5 built using the same number of HDs.
>And a RAID 6 will be more robust than either. Basing reliability on
>"hopefully you wont have both disks in a mirror fail" is just silly.
>Either you need double disk failure protection or you don't.
That statement is incorrect and ignores both
probability and real world statistical failure patterns.
The odds of a RAID 10 array of n HDs suffering a
failure that loses data are less than the odds of
it happening in a RAID 6 array of n HDs. You are
correct that RAID 6 is more robust than RAID 5.
In response to
pgsql-performance by date
|Next:||From: Arnau||Date: 2005-12-29 09:48:26|
|Subject: How import big amounts of data?|
|Previous:||From: Michael Stone||Date: 2005-12-27 21:15:18|
|Subject: Re: What's the best hardver for PostgreSQL 8.1?|