| Title: | DEC Rdb against the World |
| Moderator: | HERON::GODFRIND |
| Created: | Fri Jun 12 1987 |
| Last Modified: | Thu Feb 23 1995 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 1348 |
| Total number of notes: | 5438 |
The following message is from an Oracle employee describing how
they get some of their performance numbers. This comes from
Usenet, a world-wide message exchanging system.
From: [email protected] (Charles Simmons)
Newsgroups: comp.arch
Subject: Re: 64-bit addresses
Date: 6 Mar 90 08:11:15 GMT
Organization: Oracle Corp
Here at Oracle, we run TP1s on lots of different platforms. Currently
the Amdahl 5990 is the fastest TP1 machine. [Hmmm... Unless Amdahl
has announced something more recent than the 5990.] A good industrial
strength mainframe can generate something on the order of 300 or so
transactions per second. [It might be 250, it might be 400, I don't
remember the exact number. But it isn't 1000, unless you cheat a lot.]
A good fast mini such as a Sequent with 20 or so processors can execute
on the order of 120 transactions per second. I think you could get
a Sun 3/50 or 3/60 to do around 10 transactions per second. For the
current generation of RISC workstations, 20-30 TPS is the figure to shoot for.
What kind of hardware do you need for high TPS rates? Fast disks.
When we're doing performance analysis of a port, we play lots of little
games like using a real small database that fits in memory, etc. On
a 5990, if your database fits in memory, and if you turn off logging
[Kids! Don't try this at home!] you can get 1000 or so TPS. But as
soon as you start doing I/O, your performance drops dramatically.
For example, the Apollo Prism (4 processors) can do around 120 TPS...as long
as you don't do any I/O. Start doing I/O and the number drops to 30 or
40. Also, EDASD helps your TPS rates a lot.
An interesting measure is cost per TPS. A mainframe costs around
$50,000 per TPS. A PC costs about $1,000 per TPS.
Currently, the various development groups around here are having a bit
of a contest to see who can put together the first machine to execute
1000 TPS (with I/O turned on). The next goal will be to do 1000 TPS
at a cost of $1,000 per TPS.
-- Chuck
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 594.1 | Oracle...California Wonderland! | DPDMAI::DAVISGB | Escapee from New Hampshire... | Mon Mar 12 1990 22:11 | 9 |
No wonder these folks don't have a sense of reality when dealing with
customers. They're in a little dream world of their own....
Must be a scary place to work...
(or exciting, depending upon your perspective!)
Gil
| |||||
| 594.2 | In fact it's twice as fast, yeah that's the ticket | SRFSUP::LANGSTON | Rdb is *the* Relational db for the VAX | Tue Mar 13 1990 01:26 | 33 |
� On a 5990, if your database fits in memory, and if you turn off logging
� [Kids! Don't try this at home!] you can get 1000 or so TPS.
Hmmm...
� Currently
� the Amdahl 5990 is the fastest TP1 machine. [Hmmm... Unless Amdahl
� has announced something more recent than the 5990.] A good industrial
� strength mainframe can generate something on the order of 300 or so
� transactions per second.
Okay... So if I wanted to make my numbers look as good as possible on
a big machine like a 5990, I could get 1000 TPS by using a small
database, but a larger database (a large database) would yield TPS
rates 1/3 that.
Tell me more.
� For example, the Apollo Prism (4 processors) can do around 120
� TPS...as long
� as you don't do any I/O. Start doing I/O and the number drops to 30 or
� 40. Also, EDASD helps your TPS rates a lot.
I see, once again, more reasonable numbers are 1/4 to 1/3 the rates
you get when you "play lots of little games."
So that means that, if I claimed, say, oh 66 tps - yeah that's a nice
round number, I would really only be able to deliver in the 16-22 tps
range, in the real world.
� But it isn't 1000, unless you cheat a lot.
Sort of speaks for it self, wouldn't you say?
| |||||