Re: 750G disk details were leaked today

get used to it, it's the wave of the future :) reminds me of when I got everybody mad because I wanted to stay with a bunch of 2GB drives instead of a few 18GB drives they wanted to give me - that was 7 or 8 years ago.
Fast forward to a couple of months ago where I was invited on a call where sales sold a 4 disk RAID5 (900 GB usable of local storage not SAN) to somebody that had 40 or 50 36GB drives. They asked if it work and I said "yes. very slowly, but it will work". I could hear the sales guys pagers go off "abort mission! abort mission!". I'm sure if they would have yanked my phone out of the wall if they had been closer. I was politely uninvited from the call. afterwards they couldn't believe it. I told them that I didn't have any details on # of users, transactions, IOPs... I used some generic numbers for the drive stats and their existing system was capable of 4000 or 5000 IOPs, while the "new" one was capable of 500 IOPs. I couldn't help but laugh until I realized it would eventually be my problem. that's one from the stack of "If you want it really bad, that's the way you will get it -- really bad".

I know there's a lot more to calculating throughput, but I had to keep it simple and keep their attention.

On Apr 21, 2006, at 03:02 PM, Paul Vallee wrote:

> Comrades:
>
> As someone who has, more times than I can count, had to suffer
> through discussions on why an array made up of three huge disks in
> RAID-5 was not going to cut it for someone's performance-intensive
> database application, I was dismayed to notice today that Seagate
> has leaked an announcement of, get this, 750G disks. At 7200 RPM. I
> wept unabashedly and at length.
>
> I think we must fight against huge disks for databases with all of
> our (admittedly quite meagre as DBAs) might and power. We will
> almost certainly lose this battle, again, obviously, but not
> without a fight. As as such, in the tradition of BAARF, I have
> launched BAHD for DBs, the Battle Against Huge Disks for Databases.
>
> Please join me in my fight against huge disks for databases. Or not.
>
> http://www.pythian.com/blogs/170/750g-disks-are-bahd-for-dbs-a-call-
> to-arms
>
> Thank you,
> Paul

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