Your PSU may not have OCP as described in the article; it may simply limit the amperage. I'm certain that 18A is not enough to run your system. You've been lucky up 'til now.

 

Well ok, but lets say ot's like that. Desn't it have 18a on each rail? If I maybe hokked everything up incorrectly, souldnt this be solved by putting less disks on each?

 

Also, i searched for some power usage tests, since that other post I saw where somebody said each hdd uses 15A seemed kind of strange (15A is *a lot* of current...), and reviews I found have between 1.x and 2.x amps tops for an hdd, so something does not seem right... 

Huh? What do you mean? Like micro failures which would cause the disk (and that specific disk...) to lose power, and yet not enough to power down the system..?

For the record I did put a multimeter to the 12 and 5 v cables during post, and there was no change to the voltage values.

That's exactly what I mean. That specific disk may be sensitive to voltage sags, which would occur when the PSU hits the limits and starts to protect itself by lowering the voltage, which is a typical reaction to a small overload. A multimeter does not have the ability to accurately map voltage over time, you need an oscilloscope to see the spikes and sags caused by overcurrent conditions.

 

Also, multi rail PSU's typically only connect 1 rail on all the drive connectors, the other rails are wired to the motherboard and video card connectors, so there is no way without cutting and soldering to allow the drives to access the other rails.

 

Your power supply could still be just fine and outputting to spec, but another component is failing and drawing more current as it fails. Drives especially pull more power as they age, because bearings wear out. Fans can fail that way too, and draw excessive power while running.

Your PSU may not have OCP as described in the article; it may simply limit the amperage. I'm certain that 18A is not enough to run your system. You've been lucky up 'til now.

 

Well ok, but lets say ot's like that. Desn't it have 18a on each rail? If I maybe hokked everything up incorrectly, souldnt this be solved by putting less disks on each?

 

Also, i searched for some power usage tests, since that other post I saw where somebody said each hdd uses 15A seemed kind of strange (15A is *a lot* of current...), and reviews I found have between 1.x and 2.x amps tops for an hdd, so something does not seem right... 

Your multi rail PSU has a single rail that powers the MB and ALL your hard drives.  The other rails are used to power graphic cards - your PCIe 6 or 8 pin connectors - and to power the CPU with the 4 or 8 pin connector.

Attach a digital oscilloscope to the all of the drives and monitor the voltage over time. HP makes some great digital recording muti-meters.

 

30W is 2.5A per disk = 15A. If the rail is also used for the MB there is not enough power. Less than 1/4 of the power is available for the drives. The PSU is great for a gaming or parallel computation machine. It is not appropriate for a file server.

 

30W at 12V = 2.5Amps. 6 disk drives X 2.5A = 15 Amps on one of the 12V rails. Each rail supplies up to 18Amps. 3 Amps overhead for anything else drawing from the rail. Find a wiring diagram for the PSU to determine exactly how the rails are wired.

Attach a digital oscilloscope to the all of the drives and monitor the voltage over time. HP makes some great digital recording muti-meters.

 

30W is 2.5A per disk = 15A. If the rail is also used for the MB there is not enough power. Less than 1/4 of the power is available for the drives. The PSU is great for a gaming or parallel computation machine. It is not appropriate for a file server.

 

30W at 12V = 2.5Amps. 6 disk drives X 2.5A = 15 Amps on one of the 12V rails. Each rail supplies up to 18Amps. 3 Amps overhead for anything else drawing from the rail. Find a wiring diagram for the PSU to determine exactly how the rails are wired.

 

Yes, I thought the math was off... Your calculation is what I got, 2.5 per disk. Also i don't have six disks, I have five at the moment, so that would be 12.5A. Shouldn't it be more than enough?  I'll look for the diagram, it's the OCZ Stealth X treme, 600W

Shouldn't it be more than enough?  I'll look for the diagram, it's the OCZ Stealth X treme, 600W

 

Yes, for 5 disks that should be enough power -- especially after they're all spun up.    If you can boot the array, and all disks are spinning, then the power requirements are even lower at that point.

 

Nevertheless, a high-capacity single-rail unit has a LOT more "headroom" ... that's why for systems with a lot of disk it's preferable to use the single-rail units.

 

Also i don't have six disks, I have five at the moment, so that would be 12.5A. Shouldn't it be more than enough?

It should be, and probably was, when everything was new. Things wear out. That PSU doesn't have much extra capacity, so anything that pulls extra current as it wears is going to stress it even more.

 

I don't know what the final cause of your issues is going to end up being, it may not be the power supply, but one thing is certain. Even when brand new and operating as designed, that power supply does not supply enough 12V current to provide a healthy margin of safety. It's like towing a heavy camper trailer with a compact car. It may work, but it sure is working it hard, and it would be much easier on the equipment to have a large truck instead of the small car.

 

You keep trying to argue that the power supply is fine, and it may be, but it isn't ideal, and may cause issues.

 

Do you really want help solving the problem, or are you trolling?

Here are three options to proceed.

 

1. Try a new disk. If this doesn't work you'll have a spare and/or be ready to grow the array.

 

2. Try a new PSU. If this doesn't work you can return the PSU. Or just leave in place because the current PSU will not support more drives.

 

3. Run memtest overnight. There's no real evidence pointing to the RAM but his is an easy test and if it is the problem nothing else is going to work anyway.

 

I'd start with the memtest. Then try number 2 because you should have a spare, in any case. Then you can replace the PSU if nothing else works. It could be the MB but that is the last thing you want to replace.

Also i don't have six disks, I have five at the moment, so that would be 12.5A. Shouldn't it be more than enough?

It should be, and probably was, when everything was new. Things wear out. That PSU doesn't have much extra capacity, so anything that pulls extra current as it wears is going to stress it even more.

 

I don't know what the final cause of your issues is going to end up being, it may not be the power supply, but one thing is certain. Even when brand new and operating as designed, that power supply does not supply enough 12V current to provide a healthy margin of safety. It's like towing a heavy camper trailer with a compact car. It may work, but it sure is working it hard, and it would be much easier on the equipment to have a large truck instead of the small car.

 

You keep trying to argue that the power supply is fine, and it may be, but it isn't ideal, and may cause issues.

 

Do you really want help solving the problem, or are you trolling?

 

No, I really do. Just trying to understand why if I have more than enough current in each rail for it to work even if all connected to the same one, why it would be an issue unless of course if like you say it degrade, wore, etc in time, and trying to pinpoint what would give away if that is in fact something that happened. 

Of course if I'm having boot issues, that would also indicate a possible power problem, since like I said, leaving it off a while it boots up fine, but if I shut it down and want to boot up again it's not working....

My hesitation is simply due to the fact that sadly in my country all these things are really expensive due to excessive taxes, exchange rates, etc. (Just so you have an idea, a "good" psu like you're recommending costs about 200-250 US dollars, and besides the nominal value it, would represent about 20-25% of a good monthly paycheck...).

 

So anyway, if it is that than well, nothing you can do, I would go and replace it, just want to be as sure as possible, since I can't afford to spend that much and then find out it's not the cause and have to look elsewhere . Hope you understand, and I really really appreciate all your help!

Here are three options to proceed.

 

1. Try a new disk. If this doesn't work you'll have a spare and/or be ready to grow the array.

 

2. Try a new PSU. If this doesn't work you can return the PSU. Or just leave in place because the current PSU will not support more drives.

 

3. Run memtest overnight. There's no real evidence pointing to the RAM but his is an easy test and if it is the problem nothing else is going to work anyway.

 

I'd start with the memtest. Then try number 2 because you should have a spare, in any case. Then you can replace the PSU if nothing else works. It could be the MB but that is the last thing you want to replace.

 

Think I will try 3 first (i did leave it running some time, but not overnight, so i wil just in case). Ten could be 1 and 2 in order. The only thing throwing me off is the boot issue I mentioned. That is new, and it may just be a clear indication that says dude, give it a rest, your psu is fu*ked up...

 

I will try to re arrange the drives on the different connectors though as well.

If you do get a new PSU, do not worry at all about wattage. Only the number of amps on the 12V rail matters. For example, the Corsair CX430 has 32A on the 12V rail.

If you do get a new PSU, do not worry at all about wattage. Only the number of amps on the 12V rail matters. For example, the Corsair CX430 has 32A on the 12V rail.

 

Ok I'llkeep this in mind.  S looking at the ratings for my psu, and trying to understand all the info you guys gave me, if I look at the label (pic attached) do you mean that of the 4 "rails" listed, 3 are for the motherboard and pci-e video etc, and for EVERYTHING else you may connect (hdds, cds, etc) you are left with just a single rail and 18A max?

 

I also found this interesting from a review site:

 

"We were happy to see that this power supply doesn't turn on if you try to pull more power than ot can deliver"

 

http://www.hardwaresecrets.com/article/OCZ-StealthXStream-600-W-Power-Supply-Review/451/8

 

Think that could be the case to regarding the boot issue?

 

 

 

 

See this page: http://www.hardwaresecrets.com/article/OCZ-StealthXStream-600-W-Power-Supply-Review/451/6

 

 

12V3 has all of the SATA and MOLEX connectors as well as the MB.  The MB alone can consume 5 amps.

See this page: http://www.hardwaresecrets.com/article/OCZ-StealthXStream-600-W-Power-Supply-Review/451/6

 

 

12V3 has all of the SATA and MOLEX connectors as well as the MB.  The MB alone can consume 5 amps.

 

Would this be a good choice?

 

http://www.newegg.com/Product/Product.aspx?Item=N82E16817139005&cm_re=corsair_650w-_-17-139-005-_-Product

 

 

See this page: http://www.hardwaresecrets.com/article/OCZ-StealthXStream-600-W-Power-Supply-Review/451/6

 

 

12V3 has all of the SATA and MOLEX connectors as well as the MB.  The MB alone can consume 5 amps.

 

Would this be a good choice?

 

http://www.newegg.com/Product/Product.aspx?Item=N82E16817139005&cm_re=corsair_650w-_-17-139-005-_-Product

 

Yes. For 15-16 drives.

Is disk2 still red-balled ?  If so, you NEED to get this rebuilt if you don't want to lose the data on disk2.

 

If you've successfully rebuilt it, then there IS a way to test whether or not your power supply is marginal ... but it requires building a new configuration WITHOUT a couple of your data drives ==> so you'll lose any possibility of rebuilding disk2 [Thus I would NOT do it if you haven't rebuilt disk2 yet, unless you're willing to simply lose all of its contents].

 

... you could create a new configuration with 2 fewer disks (disconnect both their data and power connections) => and see if that runs reliably for a few days.  If so, it's VERY likely your issue is power.

 

Another thing you could do if you're electronically "handy" is to rewire one of the power buses (e.g. the PCIe auxiliary power feeds, which you almost certainly aren't using) to "feed" 3 or 4 of the disks.    This would distribute the load, and would almost certainly ensure power is not an issue.

 

See this page: http://www.hardwaresecrets.com/article/OCZ-StealthXStream-600-W-Power-Supply-Review/451/6

 

 

12V3 has all of the SATA and MOLEX connectors as well as the MB.  The MB alone can consume 5 amps.

 

Would this be a good choice?

 

http://www.newegg.com/Product/Product.aspx?Item=N82E16817139005&cm_re=corsair_650w-_-17-139-005-_-Product

 

Yes. For 15-16 drives.

 

Ok, well my case has space for up to 12, so I think I would be covered . Trying to find it here locally, else another option would be to wait a couple of months and get it in the US, since I'm supposed to be going over there by end of august...

Is disk2 still red-balled ?  If so, you NEED to get this rebuilt if you don't want to lose the data on disk2.

 

If you've successfully rebuilt it, then there IS a way to test whether or not your power supply is marginal ... but it requires building a new configuration WITHOUT a couple of your data drives ==> so you'll lose any possibility of rebuilding disk2 [Thus I would NOT do it if you haven't rebuilt disk2 yet, unless you're willing to simply lose all of its contents].

 

... you could create a new configuration with 2 fewer disks (disconnect both their data and power connections) => and see if that runs reliably for a few days.  If so, it's VERY likely your issue is power.

 

Another thing you could do if you're electronically "handy" is to rewire one of the power buses (e.g. the PCIe auxiliary power feeds, which you almost certainly aren't using) to "feed" 3 or 4 of the disks.    This would distribute the load, and would almost certainly ensure power is not an issue.

 

Yes, the disk is still like that, as the attempts to rebuild have failed (most I got was reaching about 20% through the rebuild process before it stops due to too many errors) 

 

What you mention I couod probaly try with some other disks? Ad not "sacrificing" (or risking the sacrifice) of the current disks.

 

I guess the other thing you mention about rewiring is also a good possibility... I'm definitely not using for example the pci-e connectors, since this machine is dedicate to unraid and using only onboard gpu... That actually would give me another 18A rail and might solve the problem if its power. S it might be a good option, specially because the psu I have is not badly rated, it's just that it wasn't the best match for this use from all I'm learning....

... specially because the psu I have is not badly rated, it's just that it wasn't the best match for this use from all I'm learning....

Correct.  Yours was designed to be used in a Gaming rig where you would have one or two graphics cards installed and the power would be dedicated just for the cards.

... specially because the psu I have is not badly rated, it's just that it wasn't the best match for this use from all I'm learning....

Correct.  Yours was designed to be used in a Gaming rig where you would have one or two graphics cards installed and the power would be dedicated just for the cards.

 

Yeah ok, so I think definitely power supply issue,since the boot problem I just ried to see what's going on, so I just disconnected all disks, and powered it on,whichit did just fine (and this is the part I don't understand why) but then after loading memtest a while just to let is boot something,I turned it pff, and after that, subsequent power ons and the computer does not boot...  Throws me off why it does boot thefirst time though, so I don't know wtf... On top of the "stealth/silent" psu's fan makes some really bad whirring noises, so that's gota be telling something as well...

T be 100% sure I'm going to try to pull another psu from a working pc that while notsuitable for disks, it will at least let me power the motherboard and try some power on/power offs etc, just to male sure it's not something else too...

Good idea to confirm it's not just the power supply -- you don't want to buy a nice new (expensive) PSU and find out it was a more significant issue.

 

Good idea to confirm it's not just the power supply -- you don't want to buy a nice new (expensive) PSU and find out it was a more significant issue.

 

True

 

So I actually tried with the other psu I mentioned, and powered off and on a couple of times, and booted up just fine. Of course not with the disks, since this psu also is not suitable for disks (less even than the other one) but no issues, when the original p su is not even booting a bare cou/ram/mobo combo, so I guess it's time to buy the psu... Don't think I'm gonna wait though to not rosk losing another disk, and might go for what I can find locally, that while not a good price it is a good one I think, the seasonic M12ii-650

The M12II-650 is an excellent choice.

Yes. Here is more info on PSU: http://lime-technology.com/forum/index.php?topic=12219.0

Ok, so thanks again to everyone, just to show it's not for nothing, I actually did go out yesterday and bought the Seasonic PSU.

 

Hooked up everything, turned on just fine, tried a couple of reboots, etc all good to, so hooked everything up and started it.

 

Unfortunately, I'm still getting the same error, so red ball, did the reconstruction process but stopped somewhere between 10-20% again, and took out the disk...

 

Is it worthwhile to paste syslog again in case one of you more experienced folks notice something, or should I just get a new disk..?