disk/filesystem benchmarking

[f0dder]

Hey everybody,

finally scraped some cash together and got me a shining new disk: a 74GB Raptor drive. Of course I now want to benchmark the drive, and I don't really know which non-pay tools are any good? I know there's a free lite version of SiSoft Sandra, but nothing happens when I try to launch it

[jgpaiva]

Great coincidence!
A few days ago, i was in the exact same situation. I found this. It doesn't present extensive information, but works very well and alowed to, at least, compare my 2 disks and my friends' disks. It was also useful to know the read and write speed of my disks

[f0dder]

Thanks, at least that's a starting point

Doesn't seem too comprehensive, and it's annoying that there isn't some "report window" to copy-paste from... but it'll do for now. Thanks! 

[dk70]

HDtach is among standard benchmarks for hds I believe http://www.simplisoftware.com/Public/index.php?request=HdTach

See what they use at Storagereview http://www.storagereview.com/ or http://www.benchmarkhq.ru/english.html?/be_hdd.html - or this review of same disk http://techreport.com/reviews/2004q1/raptor-wd740gd/index.x?pg=1 though you probably got new 16mb cache model.

[f0dder]

I eventually re-remembered HD-tach, which is pretty nice - free version doesn't do write tests, but read test is nice enough to get a good indication of drive speed anyway. Yup, got the 16meg model.

Performance does seem decent:


But getting Windows installed on the drive always ends up with "Cannot load operating system" - *grmbl*. Somehow mixing SATA (the raptor) and PATA (my plextor DVD) always ends up giving trouble; funny enough RAID seems less problematic, probably because it shows up as a SCSI drive.

[dk70]

Hmm I thought HDtach went freeware and not limited for private use. Nice to to get some numbers since I have same disk waiting to be installed. I also have Plextor dvd so we will see. I plan to use nlite for mixing in Nvidia SATA drivers - not doing that is what bugs installation perhaps?

Faster access is what should make the big difference - for those noticing at least 

[f0dder]

Well, HD-tach is freeware for personal use only - you gotta pay if you use it for magazine reviews/whatever. And you have to pay extra if you want write tests.

I think the error has something to do with BIOSes getting confused when there's both PATA and SATA drives in the system, especially if you have your PATA drive as primary master, and SATA drive in port #1. I had a LOT of trouble with my dads computer with sata disk + pata dvd as well.

I use nLite as well, and integrating their drivers - I had to do that when I installed to my SATA RAID mirror, for obvious reasons (I don't have a floppy drive, and seeing your mirror as two individual disks isn't very helpful ).

I'm going to try moving the plextor to primary slave or secondary master, see if that helps...

And yep, access time is what's going to be most beneficial - though extra linear read speed isn't bad either

[f0dder]

Okay, I got it working. It wasn't MBR or partition-boot or active/bootable flags or anything like that. Instead, I had to change the drive access mode from "auto" to "large" in the BIOS settings (and those are the only two settings, btw) - then everything worked like a charm. Sucky BIOS

On a side note: don't "safely remove hardware" on a drive that's part of a raid array. Should go without saying, but I was curious - good thing the raid is a mirror, not a stripe... otherwise "rebuilding array" would be a lot less fun...

[f0dder]

Holy zoooooming christ!

With the old RAID Mirror, VS2005 took ~10-20 seconds to start up. Running from the Raptor drive, that's down to 2 seconds!!!

Both are timings from freshly defragged partitions... of course the raptor does have an (unfair) advantage in that windows is freshly installed without too much crud. Timing from the RAID was without any antivirus software installed, though.

I'll have to try again once I've got the system more loaded, but things do seem sweeeeeeeeeet so far! And the drive is really silent, I was afraid of 10,000rpm but appearantly that's no problem. It's (a good deal) more noisy when there's heavy head movement (as in very random seek patterns, copying a non-fragmented 4gig file doesn't cause this), but that's not a problem to me; besides I've had much noisier drives in the past, and it's probably fixable with "acoustic management" anyway.

Yay, very good purchase so far!

[nontroppo]

Have you tried DiskBench?

http://nodesoft.com/DiskBench/Default.aspx

[f0dder]

Nope, might check it out a bit later. Seems like a tool you need to take a lot of care when using - and I'm not sure it would be very representive in my case (you really need an OS on another physical disk (not just partition) than the disk/partition you want to test).