Problems with RAID 5 on HPT366

Peripherals, parts, data storage...
Post Reply
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Problems with RAID 5 on HPT366

Post by seaofcrisis »

I have a BP6 with 2 celeron 400s and I run Windows 2000 advanced server on it.

For the past two and a half years it has been a file server. It had a hard drive on IDE0 for the operating system, and two 120G drives on the HPT366 using windows' software raid 1 (that's mirroring because I don't want to lose any data!). This machine always had incredible uptime. The last time I had to turn it off was when hurricane Ivan hit and I had to evacuate. It's given me no problems at all and I've been very, very happy with it.

This weekend, I bought two more 120G drives and intended to now use software raid 5. I didn't upgrade the bios because it's always worked fine for me. Well, it locked up twice when I tried to create the volume. So, I decided to just go ahead and reinstall windows. After that, I recreated the raid volume and waited about 3 hours while it formatted the drives. It locked up again. This time, it only got to 14% completion (and that took four hours). After a reboot, disk manager said it needed to regenerate the drives. That took six hours. Then it did another format. That took at least eight hours. I eventually just went to bed. When I got up this morning though, it was done and disk manager said the drive was health. But as soon as I started copying files to it, it looked up again, and when I rebooted it wanted to regenerate and format again.

Am I doing something wrong? I hope someone can help me here. I feel like my baby is sick.
Derek
Site Admin
Posts: 2489
Joined: Tue Jul 23, 2002 3:55 pm
Location: Canada
Contact:

Post by Derek »

Hi sea! Welcome to BP6.Com.

What size is the power supply? It could be that the two extra drives are straining the power supply and depriving the motherboard of clean, stable power.

Also check for bulging, burnt or leaking capacitors. They will also cause severe stability problems.

Derek
-Derek
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Post by seaofcrisis »

thanks for the tips, I'll check that when I get home. I hope I can make it work!
Dave Rave
G'Day Mate!
Posts: 894
Joined: Thu Sep 26, 2002 4:28 am
Location: Sydney Australia
Contact:

Post by Dave Rave »

cooling fans for large hdd's are always good.
if they get hot, they don't work

also, don't use the hpt366,
what, you haven't read that anywhere here yet ?
get a plug in card, and don't buy a cheapy.
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Post by seaofcrisis »

ok, I went home for lunch and examined the mb. I do not see any damaged components, and like I said, it ran fine for more than two years. So maybe that's not the problem.

I have a 300W power supply. If it will help, I certainly don't mind buying a more powerful one.

I don't think cooling is the problem. There are 10 fans in this thing, and anyway, the HDs are not even installed yet. I have them bolted together on a custom bracket I made. There is 1/4 inch of space between each one and they are presently sitting on the table next to the computer.

I've read what other's here have said about the 366, but look, I've been using it without any problems for more than 2 years! I can't believe it broke because I plugged in two more drives! Besides, a controller on a PCI card, or (god forbid) a hardware RAID card will probably cost more money than the whole rest of the computer at this point. I swear, this thing worked like a dream until last friday!

While home for lunch, I flashed the bios to the latest one found on this site. Naturally, windows wouldn't boot at that point. lol. I guess I'll start a reinstall tonight.

thanks again for the help. Any further advice will be much appreciated.
Dave Rave
G'Day Mate!
Posts: 894
Joined: Thu Sep 26, 2002 4:28 am
Location: Sydney Australia
Contact:

Post by Dave Rave »

there are stories of when it breaks, it will destroy your dirves in the process. it's your money.

10 fans plus 120gb hdds plus 300w power supply.....
thinx i might get a bigger / better one. not sure of the draw on the voltages, but extra will need more
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Post by seaofcrisis »

continued thanks for all the help.

Last night I bought a larger, 500W power supply and on a whim I also got a PCI HD controller. It's a highpoint 133. I don't know if I like this company anymore, but I saw it on the shelf and remembered the advice here and went ahead and picked it up. It was the only brand they had in the store.

I flashed the BP6 bios to the RV one without the onboard HPT controller. Somehow, windows still sees the old controller. Is that normal?

Anyway, it was about 10:00 when I finally got everything set up and tried for the 100th time to create the raid volume. I watched it for a couple of hours and then went to bed. This morning, it was stuck on 5% formatted. I just turned it off and came in to work.

God, I'll be pissed if the onboard controller messed up those drives. That would be a major financial loss. I just can't believe it though. It worked for more than two years with no trouble whatsoever.

Is there a way to do a low-level test on a drive to see if it's still good?
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Post by seaofcrisis »

oops double post.
Dave Rave
G'Day Mate!
Posts: 894
Joined: Thu Sep 26, 2002 4:28 am
Location: Sydney Australia
Contact:

Post by Dave Rave »

get HDD Health and see if the SMART specs of the drive are kaput
if the drive is two years old it may have had enough anyway. (stupid hdd)
Dave Rave
G'Day Mate!
Posts: 894
Joined: Thu Sep 26, 2002 4:28 am
Location: Sydney Australia
Contact:

Post by Dave Rave »

also, hdd manufacturers have diagnostic software to look at the SMART specs too. check around the support links on the web sites
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Post by seaofcrisis »

Do you think it's worth my time to try a low-level format and then start over? thanks.
wickie
Posts: 12
Joined: Sat Nov 06, 2004 5:03 pm
Location: Austria

Post by wickie »

let me guess: you are using two two years old drives and two new ones?

is it the same model?
do they all support UDMA133?
is the transfer rate the same?

test each single drive on the UDMA33 port, if it can be partitoned and formatted (one after the other)

when all drives are ok, it can be a timing problem but i have no idea how to figure out.
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Post by seaofcrisis »

"let me guess: you are using two two years old drives and two new ones? "

yes, but they are on different channels. I thought that would minimize timming errors.
wickie
Posts: 12
Joined: Sat Nov 06, 2004 5:03 pm
Location: Austria

Post by wickie »

you can't be sure

I only have built RAID's with harddrives of the same model, and sometimes it's a good idea to buy 5 drives if you need 4 to have a replacement part with exactly the same specification to avoid problems like incompatibility

in your case (and when all drives are still ok) you should try the following configuration: RAID 0+1 - means two disks as stripe set and the two others as stripe set mirror

which will give you 240 GIG, double write speed, more than double read speed and increases data safety (compared to a single drive)

Stripe 0: new drive (primary master) + old drive (primary slave)
Stripe 1 (the Mirror): new drive (secondary master) + old drive (secondary slave)

or

Stripe 0: both new drives (primary channel)
Stripe 1 (the Mirror): both old drives (secondary channel)


with 4 drives of the same type RAID 5 would be the better choise (but would be a bit slower maybe)
seaofcrisis
Posts: 8
Joined: Mon Jan 10, 2005 9:28 am

Post by seaofcrisis »

Success! I'm up and running again at last!

Unfortunately, I violated the cardinal rule of troubleshooting by changing two things at once, so I can't say for sure which change fixed my problems. First, I put one new drive and one old drive on each IDE channel. The idea was that this would slow both drives to the speed of the older drive, essentially giving me four drives at the same speed.

Next I split the dynamic volumes into three parts and created and formatted each separately. This was to reduce the size of all that overhead stuff like file allocations and also to get each volume below 120Gig.

The bottom line is, it's working well now. I have all my precious data restored and once again protected by fault tolerance.

Thanks one more time for the help - this website was invaluable.
Post Reply