I have a Epox K7A+, 75GXP 45GB and new 60GXP 60GB in a raid0 array.. now, when I created the array it imediately said 'striping' in the bios utility.. wasnt sure if I had to wait for it or what, so I rebooted and ran xp setup and it saw the 90gb partition. I installed XP fine. ran hd tach and got horrible 20mbps transfers, so I got the latest hotpoint drivers from epox and the raid utility and loaded. now I get like 35mbps which is better, but only the performace of 1 drive. when I run the utility, it lists the raid controller and channel 1 and 2 but there is nothign underneith them and I cant get a status on the array. the hard drive light has been flashing every 1 second for several hours now. how long does it take to stripe? will the speed go up only after its completely done? did I do somehting wrong? plz help. thanx
Highpoint HPT370 RAID0 question
- Thread starter AlphaFox
- Start date
Lars_Coleman
Distinguished
- Feb 9, 2001
- 1,020
- 0
- 19,280
You shouldn't have to wait for striping to take place. As long as you defined the array correctly in the BIOS.
There are so many things that take effect that could cause your performance hit. The first thing I would do is take a look at your stripe size and make sure you have the latest chipset drivers for that motherboard. Your stripe size has a lot to do with performance. Maybe you would want to familiarize your self with it by reading this article from <A HREF="http://www.anandtech.com/storage/showdoc.html?i=1491" target="_new"><font color=blue>AnandTech</font color=blue></A>. That could give you general information about strip sizes and what role they play in performance with a RAID 0 array.
Other then that make sure the BIOS from the motherboard and the RAID controller are up to par. Along with the drivers for the RAID controller and the chipset on the motherboard.
<font color=red>1GHz AMD x MSI K7T-Turbo x 512MB PC133 x 2-Maxtor 30GB/RAID 0 = Stream Line Butterfly</font color=red>
There are so many things that take effect that could cause your performance hit. The first thing I would do is take a look at your stripe size and make sure you have the latest chipset drivers for that motherboard. Your stripe size has a lot to do with performance. Maybe you would want to familiarize your self with it by reading this article from <A HREF="http://www.anandtech.com/storage/showdoc.html?i=1491" target="_new"><font color=blue>AnandTech</font color=blue></A>. That could give you general information about strip sizes and what role they play in performance with a RAID 0 array.
Other then that make sure the BIOS from the motherboard and the RAID controller are up to par. Along with the drivers for the RAID controller and the chipset on the motherboard.
<font color=red>1GHz AMD x MSI K7T-Turbo x 512MB PC133 x 2-Maxtor 30GB/RAID 0 = Stream Line Butterfly</font color=red>
I flashed the bios to the latest level before creating the raid. I read that article before I created the array also. the strip size is 64K, the largest you can make and the best performing according to the anandtech article. I am using the latest raid driver from eopx. I still only get 35+ performace out of it. it spikes to 60 in the beginning, but bottoms out at 35 the entire time. what do you get for performance? is there a better benchmark?
Lars_Coleman
Distinguished
- Feb 9, 2001
- 1,020
- 0
- 19,280
I'm running two Maxtor 30's at ATA100 7200RPM, and I get about 43MB/sec sustained. I use HDTach. I feel that is a worthy bench mark utility. But I'm also using a Promise chipset and not HighPoint. I haven't heard good things about HighPoint.
Another thing is that I only seem to get a good bench under Windows 98 & Millennium. I tried it on 2000 and it was a dog to. Who knows. I'm running XP now and haven't bothered to run it again ...
You know, I wasn't paying attention to the minor detail that you weren't running the same drives in the array. On IBM's site the specs of the drives are basically identical, but the 60GXP has a higher sustained transfer. I think the only reason is because of the track density and the areal desity of the platters on the drive. The tracks and platters in general hold more data, and therefor the heads don't have to move as much to get the data. That's the only reasoning I can come up with (seeing as they have about the same specs for each drive) that it would have a faster transfer rate. Most RAID controllers suggest using at least the same drive with the same specifications. You have the same specifications, but one transfers faster then the other and that could be what's messing your sustained transfer as a whole.
<font color=red>1GHz AMD x MSI K7T-Turbo x 512MB PC133 x 2-Maxtor 30GB/RAID 0 = Stream Line Butterfly</font color=red>
Another thing is that I only seem to get a good bench under Windows 98 & Millennium. I tried it on 2000 and it was a dog to. Who knows. I'm running XP now and haven't bothered to run it again ...
You know, I wasn't paying attention to the minor detail that you weren't running the same drives in the array. On IBM's site the specs of the drives are basically identical, but the 60GXP has a higher sustained transfer. I think the only reason is because of the track density and the areal desity of the platters on the drive. The tracks and platters in general hold more data, and therefor the heads don't have to move as much to get the data. That's the only reasoning I can come up with (seeing as they have about the same specs for each drive) that it would have a faster transfer rate. Most RAID controllers suggest using at least the same drive with the same specifications. You have the same specifications, but one transfers faster then the other and that could be what's messing your sustained transfer as a whole.
<font color=red>1GHz AMD x MSI K7T-Turbo x 512MB PC133 x 2-Maxtor 30GB/RAID 0 = Stream Line Butterfly</font color=red>
without the latest highpoint driver I got like 20meg transfer. ack! but with it it gets 35 average over the entire drive, according to hd tach. it seems much faster tho than just 1 drive, so I think that the bench isnt doing it justice. in sandra it scored slight less than the highest benchmark which was a ata100 30gb raid0 drive. it spikes to 60 in the beginning of the drive with hd tach and its basicl a solid red blob from 60 to 35, so I think you are right about the 2 drivers being slightly different. it must be throwing off the benchmark or something, the system seems much faster.
Facebook
Twitter
Instagram