Thank you very much for sharing.
P.V.Anthony
On this day, 03-August-2007 1:41 AM, Steve Herber wrote:
The point of Unix has always been to manage resources. In the old days,
where you only had a single CPU, the kernel had to time share among all
the different processes. The ps command shows you all the different
processes. Your system will have many processes.
Once you go to a multiple-cpu system you actually get to take advantage
of the original Unix choice of fork/exec to create new processes. Each
process can run on a different cpu core. Now instead of timesharing a
single core across every process, you can have 2 or 4 processes running
on their own core.
In you sample below, I could see apache running on one core, mysql on
another, postgres on another, while the applications written in perl,
ruby, and php running on the 4th. Of course you will really have
hundreds of processes running at a time so all 4 cores will have to
timeshare. As others said, you don't get 100% from each new core.
There is overhead from the kernel to figure out where to put a process
for the next slice and there will be more memory contention because you
have 4 CPU's talking to memory. A 4 core system will have almost twice
the cpu power of a 2 core system. Most of your processes look like they
will be IO bound but when a CPU bound tasks runs, it will only take over
one core leaving the other three to handle the IO bound ones.
Your initial issue is cost and rightly so. A 2 core system is cheaper
than a 4 core system. If you build equivalent 2 core and 4 core systems
then you could test them against you real workload and see if the 2 core
system is better. You might discover that you should really have 2 two
core systems for the load you have and for redundancy. Or you might
discover that you really need to address the IO bottleneck and get
multiple SCSI data paths. I would buy a 4 core system with lots of
disks. Then I would run a test to benchmark my system, probably just
running top to see the 15 minute load average. Then I would use the cpu
control software to disable one, two, and then three cores while my
normal load was running and I would again look at the 15 minute load
average. I would expect to see it rise as the number of CPU's was
reduced. My next computer would have the number of cores that met the
sweet spot.
Cheers!
Steve Herber [EMAIL PROTECTED] work: 206-221-7262
Security Engineer, UW Medicine, IT Services home: 425-454-2399
On Thu, 2 Aug 2007, P.V.Anthony wrote:
On this day, 02-August-2007 4:58 PM, Pascal BERTIN wrote:
P.V.Anthony a écrit :
<snip/>
> Apps.
> 1. Gentoo linux 64bit
> 2. Apache 2
> 3. MySql
> 4. Postgres
> 5. Qmail
> 6. Pure-ftpd
> 7. Mod_perl
> 8. php
> 9. ruby
>
<snip/>
aren't you forgetting the app that you will run the most :
emerge.
This one will really benefit from quad core, and during time where you
will work on the server.
So, facing such a choice, I would also consider my comfort/time, and
give quad core a +1.
Pascal
What I am really interested is, how the server will perform most of
the time. Will the instance of the apps go to each core in a balanced
way?
From the previous posts, I gather that the amd chip is really good.
I must say that I love the intel drivers for the network and sata. If
only amd has some chip sets for their own cpus and good open source
drivers for their chip sets. Intel has that covered and I think the
intel drivers are open source.
Please correct me if I am wrong.
P.V.Anthony
--
[EMAIL PROTECTED] mailing list
!DSPAM:6,46b21781197995167217508!
--
[EMAIL PROTECTED] mailing list
