Server won't boot after parity drive replacement

[ElJimador]

Here's a mystery for anyone who can help.  I'm starting a project to optimize for Plex (ie. Handbrake) all the full rip blurays that I keep on my shared server (listed as the main server in my sig) and as I do that, harvest the 6TB drives I no longer need there to swap them out for the 3TB drives on my home server (the backup server in my sig) to increase capacity where I need it.  Step 1 yesterday was to replace the parity drive.  Simple operation.  Except once I swapped the drives out (plus a couple case fans) the server wouldn't power up.  First I figured I just forgot to turn the PSU on before I closed the case or maybe I knocked loose the power switch wire from the front panel connections on the motherboard.  However when I popped the hood I saw that neither was the case.  So take it into my lab and with the lids still off I try again, only this time it boots on it's own as soon as I plug in the power cord.  Weird but OK I guess.  Check the unRAID GUI from my laptop, see the old parity drive missing and the new 6TB available as the replacement and all other drives present.  So all good right?  Until I powered down and tried to turn it back on again.  Nothing.  Now I'm convinced that the problem must be with the power button, that maybe the precarious looking wire lead to it got yanked off at the front panel.  So even though it's not obviously damaged or hanging loose I go ahead and swap out the front panel from the twin Node 304 that I use side by side with it to house the extra drives connected via the sata controller.  But then it still wouldn't power on after that.  So now I start following the usual troubleshooting I'd do for boot problems with a new build: double checking every connection, re-seating the RAM, removing the sata controller from the PCI slot and disconnecting the drives connected onboard, etc.  Can't remember the exact order I tried all these things but long story short, I was never able to get it to boot up again with all 12 drives attached.  The best I could do was get it to boot with just the 6 drives connected onboard attached but even that stopped working the longer I fought with it.  Along the way I'm trying to change BIOS settings whenever I can get a slimmed down configuration to boot, in between clearing RTC and CMOS (by removing the battery) when I can't, all to diminishing returns.  By the time I quit with it, it wouldn't boot except with no drives attached and the last time it wouldn't output any video (??).

 

But wait, there's more.  I happen to have another ITX computer I'm re-purposing for my sister, so to rule out a hardware issue I swap out the PSU from that rig and it still won't boot (w/only the drives connected onboard attached).  Then finally I go back to the PSU I'd been using and do a complete rebuild with that + the other ITX board (ASUS H87i-Plus w/a Pentium processor).  First attempt to boot with just the drives connected onboard worked, so I'm thinking I'm making progress.  But when I put the sata controller in the PCI slot it wouldn't power up again even with no drives attached to it.  (As with all other failed boot attempts, the only signs of life were that the fans flinch like they're trying to start and if I have the sata card in the slot the light on it will flash too.) So of course I start doing the same alternating between changing settings in BIOS and clearing CMOS, hoping somewhere I'm going to stumble on the magic combination that's going to make it work again.  (And no doubt probably only mucking things up even further in the process).

 

Again, there's no way I can remember the sequence of every step I tried in what order because it turned out to be the entire freaking day I fought with this, but the only theories I was left with by the end of all of it were: 

 

1) I changed something early on in BIOS (or w/the replacement board had the wrong settings from the start) without realizing their importance -- particularly related to the PCI or graphics output, since I don't recall if I ever got the original board to boot again with the sata controller back in the slot either.  

 

2) I drained the motherboard batteries (or otherwise damaged anything else that could be?) by clearing CMOS too many times. (Yes on both computers, because I'm that stubborn and hours of beating my head against a wall left me w/no better ideas).  

 

So with apologies for the long post, I come here hoping someone can give me any suggestions on what might have happened or what I can do at this point to try to get it working again with the original motherboard and PSU (now that I at least know that neither of those was the problem to begin with).  

 

Anyone?  A zillion karma points to any good soul who can help me untangle this mess.  Thanks!

 

[SSD]

Is the motherboard somewhat old. Some motherboards use capacitors that swell and intermittently fail. I had a tv that suffered from this. At first, repeated attempts to power on were successful. And it would run forever, but eventually I could not power it on. It was a known problem with this tv, and even out of warranty I got Samsung to repair it. Took the guy 5 minutes to replace three tiny capacitors.

 

I also had a motherboard that failed due to the capacitors. ASUS P5W-DH Deluxe. No such luck on repair, and I replaced it with a model advertising better capacitors.

 

Not sure this is your problem, but could be.

[Frank1940]

29 minutes ago, bjp999 said:

Some motherboards use capacitors that swell and intermittently fail.

This is easy to check for.  Look at all of the capacitors and see if any have a bulged top.  That is the tell-tale sign of a capacitor with the issue.  (It was a manufacturing defect...) 

[ElJimador]

8 minutes ago, bjp999 said:

Is the motherboard somewhat old. Some motherboards use capacitors that swell and intermittently fail...

 

Thanks BJP.  I was thinking it might be some defect w/the motherboard as well until I swapped the board out and still couldn't power up once I placed the sata controller in the PCI slot.  To answer your question, I bought the board I'd been using all along (ASUS C60M1-i) in 2013 and have had it in pretty much 24/7 operation running unRAID ever since (though I have power cycled occasionally including 2 or 3 times over the past several weeks for moves and to add or replace failed drives, never with any problems booting back up again).

 

With the original board, I am concerned that I may have ultimately damaged something myself by repeatedly pulling the battery to clear CMOS in between failed boot attempts.  (Is it possible to kill the battery or otherwise damage the board itself in this way?)  As I said, that suspicion is based on the diminishing returns on boot attempts the longer I struggled with it (ie. being able to boot only with onboard drives attached, then no drives, and finally not being able to get it to output any video).  Other than anything I could have done to it myself though I really doubt there was anything wrong with it to begin with. 

[ElJimador]

5 minutes ago, Frank1940 said:

This is easy to check for.  Look at all of the capacitors and see if any have a bulged top.  That is the tell-tale sign of a capacitor with the issue.  (It was a manufacturing defect...) 

 

Good tip, thanks.  And no, none of the capacitors look bulged to me.

[Frank1940]

13 minutes ago, ElJimador said:

 I was thinking it might be some defect w/the motherboard as well until I swapped the board out and still couldn't power up once I placed the sata controller in the PCI slot.  

 

Could it be something on that SATA controller?  It almost sounds like the PS(s) is(are) detecting a overload condition and that will cause a shut down of the PS.

[ElJimador]

2 hours ago, Frank1940 said:

 

Could it be something on that SATA controller?  It almost sounds like the PS(s) is(are) detecting a overload condition and that will cause a shut down of the PS.

Yeah I definitely suspect something amiss here since with both boards I was originally able to boot with all onboard drives attached but not with the sata controller installed.  But this is where I'm wondering if it's more likely something I screwed up in BIOS than a problem with the card itself.  Because it's been a long time but I vaguely remember issues with both boards when I first tried to install a sata controller in the PCI slot.  Can't remember if it wouldn't power on at all or if it just wouldn't recognize the drives attached but I seem to recall having to change something in BIOS to get the card to work (think maybe something under the PCI or graphics output though my memory at this point is so faint I couldn't really say).  And btw the card was already flashed to IT mode when I bought it so it wasn't a driver issue or anything like that.

 

Isn't there something with most ITX boards that if you use the PCIe slot for anything other than a graphics card that you might need to change something in BIOS to get it to work?  (Or for that matter w/larger boards too if you use the primary slot intended for a graphics card?)  Whatever the settings need to be, it doesn't seem like it should fail to boot entirely however I can't help thinking that I unintentionally changed something there (on the original board) or had it set incorrectly to begin with (on the replacement) since it's about the only thing I could think of that would explain why I couldn't boot the full array with either motherboard.  

 

And yeah I suppose it could be the card itself as opposed to anything I did in BIOS.  I do have another SATA controller I can try, I'm just reluctant to use it since I already pulled it from my other unRAID server after I had issues with drives mysteriously dropping off (though I've since learned it was more likely just a bad SAS cable).  Maybe I'll have to try it though since that and the RAM are the only components I haven't tried swapping out yet.

[Frank1940]

It seems to me that there were issues with using non-video cards in slots which the MB manufacturers had designated as for video cards.  There was usually some place to change that somewhere in the BIOS settings.  (You might have to wonder through all of the BIOS screens to find it.)

 

One other thing to consider is that the MB battery has gone dead. This could explain why things went south after you removed the power from everything by pulling the plug...   Those batteries are usually cheap and can be found about anywhere they sell batteries for hearing aids. 

[ElJimador]

3 hours ago, Frank1940 said:

It seems to me that there were issues with using non-video cards in slots which the MB manufacturers had designated as for video cards.  There was usually some place to change that somewhere in the BIOS settings.  (You might have to wonder through all of the BIOS screens to find it.)

 

One other thing to consider is that the MB battery has gone dead. This could explain why things went south after you removed the power from everything by pulling the plug...   Those batteries are usually cheap and can be found about anywhere they sell batteries for hearing aids. 

 

Thanks Frank.  I was wondering whether the motherboard battery is replaceable.  It looks standard enough but I wasn't sure if it might be something proprietary (odd voltage or whatever) that you could only get replaced by the board manufacturer.  If that's the case, my dad's got an old Simpson meter I can use to test it and get a replacement if necessary.  That's a great suggestion.

 

As for the BIOS settings to get the sata controller to work, I seem to recall it's something really counter-intuitive like you have to set the graphics output to PCI or enable multi-monitor even though you're using single output integrated graphics.  Makes no sense but it was something like that or some combination things that I never would have tried because on their face they're just wrong yet somehow it works anyway.  Damn I wish I could remember what it was because I feel like I tried every possible combination already.  Of course if I did finally stumble on the right combo but only after I drained the battery then it would have been a moot point by then anyway.

 

So alright, I've got some avenues to work on here.   Thanks so much for your input and helping me think through this.  I'll report back after I've had a chance to test/replace the battery, try the various BIOS settings, and if still necessary try the other SATA controller too.  If you have any other suggestions in the meantime please keep them coming.

[ElJimador]

Well trying again didn't take very long.  Rebuilt with the original board and was able to boot with no drives attached and set BIOS to what I hoped were the right settings.  (Got my notepad to document each combo on the NB settings until I hit the right one.)  Next installed the sata controller with a single SSD attached to it, no dice.  Won't power up.  So I take the card out so I can go into BIOS again and try the next combo and it still won't power up.  With no PCI card and no drives attached.  And BTW the battery was fine.  Checked it with the Simpson meter and it's still at full voltage.  

 

So I guess I'm shopping for a new motherboard.  Shame really.  I loved that board and last time I looked you couldn't find anything like it anymore.  (Fanless 9w TDP processor on an ITX board with 6 sata6 ports).  Oh well. 

 

Thanks again though Frank.  I really do appreciate the help even though it didn't work out.

[ElJimador]

Okay, I thought I had at least diagnosed the problem (bad board) and was done with this but...  no.  Today I decided I was going to rebuild with the replacement board I tried earlier (ASUS H87i-Plus w/Pentium G3258) and just use that for my home server going forward.  And guess what?  Exactly the same result as yesterday with the C60M1-i.  First boot with nothing attached but the fans and the unRAID thumb drive works and in BIOS I set the system date/time, change the fan profile to Silent, and in the NB settings I select primary display = iGPU and multi monitor = enabled (I finally found an old post here that said that was the trick w/the C60M1-i to get the SATA controller to work).  That's it though.  Every other setting I left at its default and I certainly didn't touch anything in the AI Tweaker section (I never do since I don't OC).  Power down, install the sata controller (the SAS2LP-MV8 this time, which I'd never used in this server earlier) and it wouldn't power on.  So fine, turn off the PSU and unplug it, pull the SATA controller and try to boot into BIOS again to try some different NB settings and IT WOULDN'T EFFING BOOT.

 

Take a few minutes, calm down, assume that I had misdiagnosed the problem and that it was really the PSU all along and swap out the PSU to see.  Now it turns on (the fans are going at least) but with no graphics output.  And as before when I got no graphics (with the original board) I can't just reboot from the unRAID GUI because I'm hitting delete at startup thinking I'm going into BIOS.  So in the blind I try hitting F10 and enter to see if I can get it to exit BIOS and boot into unRAID but no dice.  So finally there's nothing to do but hard power down.

 

Now seriously, WTF could possibly be going on???  2 entirely different everything now -- motherboard, PSU, RAM, sata controller (when I can even get to a 2nd boot attempt) -- both using 99% default BIOS settings and I still can't get either one to power up.  And there's been no static electricity discharge or anything like that either.  I wear the wrist thingy and am always careful about that.  Not to mention every component was just pulled from a perfectly working system.

 

Does anyone else have any ideas??  Please?  Because I'm feeling real close now to re-enacting the printer scene from Office Space and taking a baseball bat to the whole freaking works if I can't start making progress on this pretty soon.  Thanks.

[SSD]

I can only suggest to try to get a basic configuration working consistently. Start with motherboard, PSU, 1 stick of RAM, video, monitor and keyboard. No hard disk, USB, no add on controller. With 2 motherboads acting up, the PSU or the RAM are looking like good candidates. Make sure the motherboard is isolated from the case. I once had a motherboard that would not boot in the case but did boot on my bench and turned out to be a fluke short in the mounting.

 

If you think it was the BIOS, you can do the shorting technique to reset the BIOS.

[ElJimador]

24 minutes ago, bjp999 said:

I can only suggest to try to get a basic configuration working consistently. Start with motherboard, PSU, 1 stick of RAM, video, monitor and keyboard. No hard disk, USB, no add on controller. With 2 motherboads acting up, the PSU or the RAM are looking like good candidates. Make sure the motherboard is isolated from the case. I once had a motherboard that would not boot in the case but did boot on my bench and turned out to be a fluke short in the mounting.

 

If you think it was the BIOS, you can do the shorting technique to reset the BIOS.

Thanks BJP.  I know my frustration is showing at this point so I want to be clear that I really do appreciate anyone willing to hang with me through these long posts and continue to help me out with this.  And yeah, now that you mention it I guess the case is the one constant component in all these struggles that I hadn't considered before (that and the unRAID thumb drive, as unlikely a culprit as that would be).  I have checked in swapping out the different parts that there wasn't anything obvious like a loose screw wedged under the motherboard creating a short.  But after swapping out everything else it probably is the most rational explanation left, and one that may not have ever occurred to me on my own.  So thank you. Tomorrow I'll try a different case and see if that finally solves it.

 

Again, thanks for the feedback and fingers crossed that things go better tomorrow.

[JonathanM]

8 hours ago, ElJimador said:

Tomorrow I'll try a different case

For testing purposes (and ONLY testing), try using a motherboard box as a stand for the motherboard, so you can hang any PCIe cards over the edge when you get that far. Start with the bare minimum, motherboard cpu PSU and see if you get missing memory beep codes. Add memory, and see if you get video, add usb boot, etc. When it gets that far, see if it will start a memtest successfully. Continue one thing at a time until things break, then back out and see if you can recover just that one step.

[ElJimador]

9 hours ago, jonathanm said:

For testing purposes (and ONLY testing), try using a motherboard box as a stand for the motherboard, so you can hang any PCIe cards over the edge when you get that far. Start with the bare minimum, motherboard cpu PSU and see if you get missing memory beep codes. Add memory, and see if you get video, add usb boot, etc. When it gets that far, see if it will start a memtest successfully. Continue one thing at a time until things break, then back out and see if you can recover just that one step.

 

Thank you Jonathan.  I followed your advice and unfortunately the results weren't pretty.

 

C60M1-I

No speaker header on board so no beep codes however everything worked fine up through installation of RAM, VGA, mouse, keyboard, unRAID flash drive, and ethernet with no problems booting into unRAID startup menu and starting memtest.  After that though:

1. install SATA controller (SAS2LP) in PCI slot = wouldn't post
2. remove SATA controller and try again = wouldn't post
3. clear RTC RAM and try again = posts, then in BIOS choose Optimized Defaults and exit and it goes on to boot into unRAID
4. power down and install different SATA controller (SASLP) = wouldn't post
5. remove SATA controller and try again = still wouldn't post

H87I-Plus

1. First try w/nothing but CPU gave me appropriate beep code for missing memory
2. w/RAM installed now (single stick only) gave me beep code that I interpreted as missing VGA (1 continuous beep followed by 3 short beeps) but in hindsight I realized was actually for hardware component failure (1 continuous beep followed by 4 short beeps) 
3. w/VGA connected got the same beep code for hardware component failure however I did get video output to a prompt to F1 into BIOS setup along with the error message that no keyboard or CPU fan were detected (might not be what you intended but I was following your steps literally and hadn't plugged it in yet, so I assumed the hardware component failure beeps on both occasions were for the missing CPU fan)
4. hard power down, attach CPU fan and try again = fan spins up and no beeps but no video either
5. clear RTC RAM and try again = same result
6. remove battery + clear RTC RAM and try again = same result
7. install video card (GT430) into PCI slot, connect VGA to it and try again = same result again

So what do you think?  My best guess at this point is that there probably was some kind of fluke short in the mounting (though I still have no idea what might have caused it) and in my repeated attempts to get things working again I inadvertently fried both boards.  I'm certainly open to any other theories though or anything else that might still be worth trying.

 

I suppose I could try ASUS support but since both boards are out of warranty that just seems like an exercise in further futility to me.  (I wound up having to RMA a monitor with them before and I can't say I found them to particularly helpful).  So unless there's some other avenue I'm not thinking of, I guess I'm buying a new motherboard??

 

 

 

 

 

[Frank1940]

Before you spring for purchase of a new MB. I would try another PSU first.  I find it hard to believe you have TWO MB's fail in a short period of time.  Plus, the C60M1 did boot until you tried to stuff a SATA controller (and two different ones at that.)  The only common thread in this whole scenario is that PSU...

[SSD]

14 minutes ago, Frank1940 said:

Before you spring for purchase of a new MB. I would try another PSU first.  I find it hard to believe you have TWO MB's fail in a short period of time.  Plus, the C60M1 did boot until you tried to stuff a SATA controller (and two different ones at that.)  The only common thread in this whole scenario is that PSU...

 

Yes, above I said that RAM and PSU are the most likely candidates. All results consistent with a lack of power. 

 

19 hours ago, bjp999 said:

I can only suggest to try to get a basic configuration working consistently. Start with motherboard, PSU, 1 stick of RAM, video, monitor and keyboard. No hard disk, USB, no add on controller. With 2 motherboads acting up, the PSU or the RAM are looking like good candidates. Make sure the motherboard is isolated from the case. I once had a motherboard that would not boot in the case but did boot on my bench and turned out to be a fluke short in the mounting.

 

If you think it was the BIOS, you can do the shorting technique to reset the BIOS.

 

Do you have a second PSU to run some tests? If not, you might drop by a mom and pop computer store (if one still exists near you) and get them to test it out. Or buy a new one from a place with a good return policy.

[ElJimador]

4 hours ago, bjp999 said:

Do you have a second PSU to run some tests? 

Yes I do and I guess just to leave no stone unturned I'll go ahead and do the same exact tests tomorrow.  But as I said in my earlier posts, I've already tried swapping out the PSU in my earlier struggles and have gotten all the same boot shenanigans.

 

And as low as the possibility is of 2 of the same kind of components (motherboard, PSU, RAM) dying at the same time, the most likely culprit would be some kind of electrical short or surge, wouldn't it?  It's not likely to jjust be sheer coincidence.  And if that happened which components would be the most vulnerable?

 

[ElJimador]

You know I was just thinking.  My "lab" as I call it is a converted front porch area of my little bungalow, so poorly insulated that I had to move my home server to the living room in summer for fear of the high temps, and I've been running without a UPS in this room since it followed the server into the living room.  I am using a surge protector power strip from Amazon but it's a $10 item.  If my landlord's contractors did as crappy a job on the wiring in this room as they did with the insulation and ventilation (also non-existant), I wonder now if I've got some kind of a dirty power problem?  

[JonathanM]

Does the surge protector have indicator lights? If so, it should tell you if the building wiring is grossly incorrect. That said, I doubt dirty power would be that consistently bad, more random crashes and such.

[ElJimador]

2 minutes ago, jonathanm said:

Does the surge protector have indicator lights? If so, it should tell you if the building wiring is grossly incorrect. That said, I doubt dirty power would be that consistently bad, more random crashes and such.

Yeah you're probably right.  I'd also probably see stuff like my monitor or the light in here flickering or that sort of thing, I can't say I've ever had any of that.  Still grasping at straws I guess.

 

BTW, check out this motherboard listing I just found on EBay.  Sounds familiar doesn't it?

 

*Broken* As-Is ASUS H87I-PLUS LGA1150 Intel H87 ITX Motherboard

System powers on, but no beep or display. Only had the i5-4460 to test the board and a set of known working DDR3 RAM. No other debug was done. The CPU Socket is clean, and the I/O shield is included.

 

 

[ElJimador]

Alright, testing again with the other PSU.

 

H87I-Plus

1. Starting RAM and CPU fan installed this time but nothing else = turns, no beeps (should be a code for missing VGA, right?)
2. Add VGA only = turns on w/no beeps or video output
3. Remove battery and clear RTC RAM = same result
4. Add graphics card and attach VGA to that, also add unRAID thumb + ethernet = still no video output and can't access the unRAID GUI through another computer either (so whatever it's doing while it's "on" without video, it's not booting into unRAID. (BTW the fan on the graphics card turns on so there is power going ot the PCI slot.)

Conclusion: Dead board just like the one on EBay.  Turning on (supposedly) but w/no video = useless.

 

C60M1-I

1. Starting where I left off yesterday w/VGA, unRAID flash, ethernet, keyboard and mouse all connected = boots where I go into BIOS and change only the NB config from the optimized defaults.  Per and old post that I finally found here, enabled multi-monitor and changed Integrated graphics from Auto to Forced while keeping primary video device as PCIe (still hoping for some magic combo that will make it work)
2. Exit BIOS and let it boot into unRAID where I power down
3. Add SATA controller (SASLP) = boots up this time.  HEY!!
4. Power down and connect SAS cable + SSD to the card = boots again and the drive is visible in the unRAID GUI.  WOOHOO!
5. Power down and go back to the original PSU with the same config = same result.  

Conclusion:  It was the *#&% BIOS settings after all.  Board appears to be okay though I still don't know how the BIOS settings got changed to start this whole mess in the first place, since I don't recall touching anything in the NB config until after it stopped booting with anything in the PCI slot.  Still, it's great news.  I will now VERY carefully reinstall it in the other Node 304 case (the one that was only housing the expansion drives) just in case there was anything in the mounting of it's twin that was creating some kind of short, and assuming it continues to work I'll start reconnecting the hard drives one at a time and then (hope, hope) start the array and move forward with the parity replace procedure.

 

Going back for a moment to the other board, I had been using it up until a few days ago in my own Windows 8.1 Pro desktop that I was in the middle of repurposing for sister's use since I was going to be upgrading to a Ryzen myself.  And unfortunately it's the OEM version of Windows with the installation media lost (tossed by my ex when we separated) AND with careless me not having saved the product key anywhere either.  So any idea in this situation if I should be able to get a new key from MS to re-install it on the replacement motherboard?  My understanding is that they will allow you to re-install the OEM version on a new board if your old board dies and I do still have the receipt from when I bought the Windows that was running on it, however there's no S/N or anything unique on that for them to be able to tie it to the original key and know that I'm not actually running it on another machine already.  

 

Damn I really wish I never even tried using that board once the C60M1-I BIOS got changed and stopped booting the PCI slot populated.  Replacing the board AND having to buy Windows again is not something I'm looking forward to (or that my sister will be looking forward to, I should say, since she'd be the one footing the bill at this point).  

 

But hey, on the bright side at least it looks like everything is going to be good again with the original board.  So thanks everyone for sticking with me through this and encouraging me to try one more time.  

[ElJimador]

And just when I thought I was going to be able to get the server back up again:

1. Install board & PSU in case and successfully boot first without the SATA controller in the PCI slot, then with it in, then w/SAS cable and SSD connected to it, and finally with all drives attached onboard connected too.  A good start.
2. Attach SAS cables to the 6 data drives in the second case and boot up where unRAID shows 3 of those drives (1,5 & 6) all missing.  Whoops.  
3. Power down and double check the connections.  One of the SAS cables (the one w D5 & 6 attached) didn't clasp in the card and had fallen out.  Not sure what happened with D1 since it did appear to be connected, but whatever.  
4. Double check that all drives attached to the controller are now firmly connected on both ends to SAS cables and PSU, disconnect all the onboard drives just to be safe, and try booting again = nothing.  Won't turn on.
5. Disconnect all drives attached to the controller and remove controller from the slot = still won't turn on.
6. Clear RTC RAM and try again = same result.
7. Remove battery and clear RTC RAM and try again = still nothing.

So now I'm back to wondering about the power supply.  I remember back when I was running this server w/only 6 drives in a single case, I was using a 300w Seasonic as the PSU for some time and at one point when I had an issue I was cautioned by someone here that the unit may not be putting out enough amperage (24a on the 12v).  Which is exactly why I bought the Cooler Master V550 when I expanded to the 2nd case and 12 drives.  550w seems like overkill to me to power a 9w TDP processor + 12 drives but I wanted to make sure I had plenty of overhead on the amps.  Specs for the V550 says it puts out max power of 45a on the 12v which if you figure 2a per drive (WD actually says 1.75a peak for the Reds which are 10 of the 12 drives I'm using but let's just round up) that still leaves 21a to power up an Atom-equivalent super low power embedded board/processor + the controller and 6 case fans.  And that's it.  So there's no way that the output should be insufficient unless the PSU is just crapping out now, right?

 

I've got another 550w PSU (Corsair RM550x) coming on monday that I was going to use in the new Ryzen desktop.  Think I'll try it with this server instead and cross my fingers that I'll be able to get it to boot up again.  The other PSU I was testing with earlier is a 400w w/only 30a max on the 12v so I don't see any point in even trying w/that since even if I could recover up to getting it to boot at all to begin with and then with up to 6 drives attached, I would never try to boot the whole array with it.  So let me know if you think there's something else I should be trying in the meantime, otherwise I'm in a holding pattern now until I get the new PSU.

 

BTW, if it does turn out that the V550 at fault and started this whole sh*tshow, does the warranty on a PSU typically cover connected components as well (like any decent UPS warranty does)?  I've never had a problem w/a PSU before but if that's really what killed the other board (and maybe took my OEM Windows license down with it) then I'd say Cooler Master owes me more than just a replacement unit.  But that's getting ahead of myself I suppose.

 

Anyway I'm calling it a day.  It's clearly time to start drinking and forget this nonsense for a while. 

 

[SSD]

I feel your pain!

[JonathanM]

5 minutes ago, bjp999 said:

I feel your pain!