Hard drive performance - how to measure

  • Thread starter Thread starter Scott
  • Start date Start date
S

Scott

Hi, I'm working on a new computer that I will mainly use for personal
stuff, but I'll also be doing some database work that requires high
storage performance.

My system is based on the EVGA 680i motherboard, and I have 4 SATA 2
drives that I can configure as either RAID5 or RAID0+1.

In order to decide which offers the best price / performance ratio -
what types of benchmarks are available to measure driver performance?

Thanks in advance!
Scott
 
Scott said:
Hi, I'm working on a new computer that I will mainly use
for personal stuff, but I'll also be doing some database
work that requires high storage performance.
Click to expand...

It shouldnt if its properly designed.
My system is based on the EVGA 680i motherboard, and I have
4 SATA 2 drives that I can configure as either RAID5 or RAID0+1.
Click to expand...
In order to decide which offers the best price / performance ratio -
what types of benchmarks are available to measure driver performance?
Click to expand...

What matters much more is the database performance.
 
Scott said:
Hi, I'm working on a new computer that I will mainly use for personal
stuff, but I'll also be doing some database work that requires high
storage performance.

My system is based on the EVGA 680i motherboard, and I have 4 SATA 2
drives that I can configure as either RAID5 or RAID0+1.

In order to decide which offers the best price / performance ratio -
what types of benchmarks are available to measure driver performance?
Click to expand...

Try HDTach ;)

Bye,
Skybuck.
 
Hi, I'm working on a new computer that I will mainly use
It shouldnt if its properly designed.
Click to expand...

Wrong if you're working on star schema dimensional models, where I
need to touch hundreds of thousands to millions of records to satisfy
a query, instead of just touching a few at a time like an OLTP system.
What matters much more is the database performance.
Click to expand...

Obviously. If I had all the free time in the world, I could test
everything thoroughly and come up with the 100% best solution. Since I
do not have all the free time in the world, I came to this group for
some helpful advice.

Since you are either unwilling or unable to provide that, why exactly
did you respond to my post?
 
Wrong if you're working on star schema dimensional models, where I
need to touch hundreds of thousands to millions of records to satisfy
a query, instead of just touching a few at a time like an OLTP system.
Click to expand...

Utterly mangled.
Obviously. If I had all the free time in the world, I could test
everything thoroughly and come up with the 100% best solution.
Since I do not have all the free time in the world, I came to
this group for some helpful advice.
Click to expand...
Since you are either unwilling or unable to provide
that, why exactly did you respond to my post?
Click to expand...

To rub your stupid pig ignorant nose in the basics, child.
 
Scott said:
Hi, I'm working on a new computer that I will mainly use for personal
stuff, but I'll also be doing some database work that requires high
storage performance.

My system is based on the EVGA 680i motherboard, and I have 4 SATA 2
drives that I can configure as either RAID5 or RAID0+1.
Click to expand...

Sysoft Sandra has a HD benchmark included. Some don't like the way it
works, but it should be fine for comparison purposes.

First guess is that RAID 10 (0+1) will be faster, at the expense of total
capacity.

Also, if you don't have 10K RPM Raptors, you're probably missing the first
benchmark: Spindle speed is important, because it determines physical
limits on access time.

BTW, ignore rodless speedless. He likes to cause problems while giving
little in the way of real information.
 
Scott said:
Wrong if you're working on star schema dimensional models, where I
need to touch hundreds of thousands to millions of records to
satisfy a query, instead of just touching a few at a time like an
OLTP system.
....
Click to expand...

Even assuming that's true, hard drive performance is relevant and
Scott's database performance might already be optimal.
Obviously. If I had all the free time in the world, I could test
everything thoroughly and come up with the 100% best solution.
Since I do not have all the free time in the world, I came to this
group for some helpful advice.
Click to expand...

Don't take it personally. Rod speed is a pathologically insulting
regular troll in whatever group he happens to be in. Sometimes he
provides good advice, but he completely refuses to accept even
gentle correction when he's wrong.
 
Scott said:
Hi, I'm working on a new computer that I will mainly use for
personal stuff, but I'll also be doing some database work that
requires high storage performance.
Click to expand...

I too would recommend 10,000 rpm drives if they're big enough. I
love my Raptor, small but fast.
 
Don't take it personally. Rod speed is a pathologically insulting
regular troll in whatever group he happens to be in. Sometimes he
provides good advice, but he completely refuses to accept even
gentle correction when he's wrong.
Click to expand...

I'll second that, he has some kind of inadequacy that he makes up for by
pretending to be a know-all in this newsgroup.

If challenged, he goes into this psycho mode of name calling, etc.

The question was why did he respond to your post, I am not sure anyone could
explain the inner workings of someone as unstable as Rod Speed.

-g
 
Scott said:
Hi, I'm working on a new computer that I will mainly use for personal
stuff, but I'll also be doing some database work that requires high
storage performance.

My system is based on the EVGA 680i motherboard, and I have 4 SATA 2
drives that I can configure as either RAID5 or RAID0+1.

In order to decide which offers the best price / performance ratio -
what types of benchmarks are available to measure driver performance?

Thanks in advance!
Scott
Click to expand...
If you are running windows HDTach has a free version.
 
Back
Top