Re: Scalable Performace - Inserts/Updates

On Thu, 21 Dec 2000 14:20:08 GMT, Brian Peasland <peasland_at_edcmail.cr.usgs.gov> wrote:

>Thanks for the correction and admission! Different vendor terminology
>can be a killer sometimes!

Tell me about it! <groan>... Sorry, once again!

>
>I'm still not convinced that going with one member for each group is the
>way to go. Personally, I like redundancy and would even go so far as to
>let Oracle multiplex the groups with multiple members at the same time
>letting hardware mirror the disk files. But then I'm not looking for
>1000 commits/second. And I know that this has been an often debated
>topic in the DBA community.
>

It has. No definite solution AFAIK. A few years ago, you wouldn't catch me without multiple members. But as hardware evolved and got much more reliable, I am now tending to single member ( ONLY in the conditions I described: EMC or similar). Where I find no RAID hardware then it's multiple members, for obvious reasons.

The diff in performance only shows in extreme environments (lots and lots of writes/commits) of course. But it's worth it.

Cheers
Nuno Souto
nsouto_at_bigpond.net.au.nospam
http://www.users.bigpond.net.au/the_Den/index.html Received on Fri Dec 22 2000 - 04:36:16 CST