Major problems with Maxtor 160G and Escalade controller

B

beamrider

Limp Gawd
Joined
Aug 26, 2000
Messages
305
Config: 8 Maxtor DiamondMax 160G drives, model # 4R160l0, all mounted in hot swap drive sleds.
3Ware Escalade model 7810.
Asus P2B-D, 512M PC133, dual PII 400's. Antec TruePower 400W power supply.

The drives are in a RAID5 configuration, with one drive a hot spare, for a total Windows formatted size of 916G.

Using the drive cables supplied with the 7810, so shouldn't be any problem with crap cables or anything like that.

I have had to RMA at least 12 of these damn drives in a 6 month period. Last Tuesday, 2 drives died. At the same time. So I do the familiar advance RMA, get the two replacements, slap them in the rig, and rebuild the array. Tonight I go down to check it, and one of the friggin replacements has died. Less than 24 hours in the machine, and it's shot. This isn't the first time it's happened like this, either.

Anyone else have any problems with anything like this? Or any suggestions for me? The drives are all still under warranty until sometime in 2005, so I don't really mind the hassle, but I know dman well once the warranty is out, drives are going to fail left and right, and I'm going to be out about $1000 in hard drives.

Could it be a fault with the RAID card itself, possibly?

Going nutz here, thanks!
 
Sometimes a controller will mark drives failed that really aren't. If you're testing the units in another machine on another controller this should tell you the truth of the matter. On the side, though, I really can't see how that PSU will power up 8 drives simultaneously; I've got a 450W Antec TP and it has issues with 6 drives and a 1.1GHz Celeron-T. Just too many amps are pulled at startup. I'm sure Ice Czar will come in later and give you links to the thread we discussed the realities of PSUs in over the past couple weeks.
 
Yeah, testing the drives on a seperate machine, all "bad" drives fail using PowerMax.

How big of a power supply do you think I should be using, then? Do they make one big enough? Or maybe I can "splice" a few PS's together, and split the drives amoungst the two.
 
Yeah, but do you really think that the PS is killing these drives?

One thing's for sure, tho, next drive I get back from RMA is going to get tested before I throw it into the machine.


And at $7 per drive to ship back, this crao is getting expensive.
 
Even with Maxtor being Maxtor, I can't see this being just the drives. The PSU certainly is a prime suspect, as is ESD & shock from having the machine messed with so often.
 
Originally posted by beamrider
Or maybe I can "splice" a few PS's together, and split the drives amoungst the two.
Click to expand...

If you end up doing this, look in the Cool Cases --> FAQ. In there is tutorial on how to wire two PSs together.

-fool
 
Originally posted by Snugglebear
I'm sure Ice Czar will come in later and give you links to the thread we discussed the realities of PSUs in over the past couple weeks.
Click to expand...

my ears where burning :p

http://www.hardforum.com/showthread.php?s=&threadid=703824&perpage=15&pagenumber=1
very long thread but with info and links to nearly my full PSU repertoire

the advantage a SCSI array would have is the option for delayed spinup

Ive been running two synchd PSUs
(that dont share a bus for over 2 years (12 HDDs)
Hardwire Synch
relay synch

there are tutorials on how to run multiple PSUs that share a bus
but Id highly advise against it
with resistors
with Mosfets
and finally how its really done with diodes
Power Supply System Integration — Part 1: Single Source Fault Tolerant Power Systems @ Power Electronics
if your going to go redundant (shared bus) buy one ;)
 
Well, on both of my other boxes, I run dual supplies with just grn and blk tied together. One PS for the drives, fans, cdroms, etc, and the other PS just powers the motherboard. Is this an OK method?

And just why in the hell doesn't this card have delayed spin up option? That's just stupid.

Taken from the spec sheet of this model of drive:

Mode 5V 12V
Seek(mA) 628 587
Idle(mA) 582 224
Standby(mA) 116 41

But they don't list spin-up surge current. Would using the seek current be OK in calculating the total max amps all 8 drives would require at once?
 
It's not the fault of 3Ware that the ATA doesn't include features like delayed spinup. You're using a bunch of drives made for individual use in desktops in a professional/enterprise capacity. They were never meant to have such features, and as such cost less for it. In my mind ATA RAID is similar to Abit's BP6, a hack to get low-end stuff to run more like high-end stuff. You can get some of the benefits by doing so, but it sure as hell ain't perfect, nor is it that close to what you're trying to emulate. Don't get me wrong, I still love my BP6 and use ATA RAID, but I don't expect them to behave like the SCSI arrays at work or the SMP workstations using P2s/P3s/Xeons/SPARCs.

As for spinup current draw, seeks aren't anywhere near what the draw is at motor start. I know my IBMs (bitch all you want but they provide more and more accurate data on their units than anyone else) suck up ~2A on the 12V rail and another ~1A on the 5V rail. Actual operating use is less than half that. This number goes up depending on the number of platters in the unit.
 
Originally posted by beamrider
Well, on both of my other boxes, I run dual supplies with just grn and blk tied together. One PS for the drives, fans, cdroms, etc, and the other PS just powers the motherboard. Is this an OK method?
Click to expand...

you should also connect the common ground (Black)
and Power Good (Grey) on the second PSU
(per the diagram in the Hardwire Synch link above)

the problem could also well be the hotswap bays
ATA Not So Frequently Asked Questions
The following article was written by snn47 to address some of the issues associated with standard ribbon cables and the use of e.g. removable drive racks as an attempt to share some insight into factors that can adversely affect the life or reliability of of desktop Hard Disk Drives. Specifically, issues like why some drives are working in some systems and not in others, the impact of cable routing and why is it that the drive manufacturers always recommend using their own cables (if supplied with the drive).

There is a ton of data in this article, some of which are of interest only for the EE-crowd or else some nerds but there is also some stuff of why sometimes simply rerouting the cable can solve the problems at hand.
Click to expand...
Id try the drives directly connected for awhile
you havent actually been hotswapping I hope
 
Originally posted by Ice Czar
you should also connect the common ground (Black)
and Power Good (Grey) on the second PSU
(per the diagram in the Hardwire Synch link above)

the problem could also well be the hotswap bays
ATA Not So Frequently Asked Questions

Id try the drives directly connected for awhile
you havent actually been hotswapping I hope
Click to expand...

LMAO, no, haven't had to do any hotswapping, I just went with sleds so it would be easier for me to switch drives in case of a failure. Box is powered down when I swap a drive out.
 
Originally posted by Ice Czar

you havent actually been hotswapping I hope
Click to expand...

No no... All the IDE cages boast hot swap.. It must be true... :D All these law suits over false advertising with HDD sizes... Someone bloody sue over false advertising on IDE hot swapping...
 
unfortuately that aint going to happen as there was limited IDE hotswap support under Win95\98 (like ejecting media from a USB card reader :rolleyes: ) and there is "true" IDE hotswap, IF the RAID card supports it and the enclosure employs a sideband technology, basically all the stuff SCSI natively has

Hot Swap Issues PDF SCSI-3
Promise Tech Sideband Hotswap
FILE: HotSwap.exe - Hot Swapping IDE or ATAPI CDROM Devices, Microsoft Knowledge Base Article - 192604

but yeh, the "hotswap" claim is widely over used and extremely misleading
 
You must log in or register to reply here.
Back
Top