My APC UPS blew up! unRAID box is dead!

[FACP]

I've been using unRAID server for a year now without any glitch.  It was running stable so I haven't touch the system for a year and never even bothered to upgrade to newer versions (was running v4.4.2).  It's been awhile since I visited the forums and read up on the latest developments.

 

Yesterday, we experienced a momentary blackout.  I have my unraid server in my walk-in closet plugged into a APC ES750 UPS.  When the power came back on, I heard a loud "pop" that came from the closet.  When I opened the door, I saw a little bit of smoke coming out of the UPS.  Oh crap!  I immediately yanked the UPS plug off the wall, ran to the kitchen and grabbed the fire extinguisher.  Whew!  Thank god it didn't cause a fire! For some unknown reason, the UPS blew up!

 

After I determined everything inside the house was safe, I went back to the closet, took the server out and tried to power it back on.  And of course, it wouldn't. Looks like the power supply and motherboard is fried. The jumpdrive with the unRAID software was toast too. I checked each data drive (6 WD 1TB Green drives), 3 were spinning up fine with all the data intact, and the other 3 (movie vault and parity) are dead!

 

Oh well, could have been worse.  I'm just thankful it didn't burn my house down.  I can always build another box and re-rip my movie collection.  I'm extremely lucky that I was home when it happened.

 

Anyway, questions...

 

 

Can I just mount the good drives and have all the data back in the array without having to copy and transfer them to a new drive?

 

Can I still use my old registration Pro Key to format another jump drive?  Or do I need to request a new one?

 

Should I upgrade to v4.5.3 or use v4.4.2?  Is 4.5.3 more stable?

 

Do you think I can get the affected drives replaced under warranty?

 

Appreciate any help or advise you good folks can give me.

[purko]

I saw a little bit of smoke coming out of the UPS.  Oh crap!  I immediately yanked the UPS plug off the wall, ran to the kitchen and grabbed the fire extinguisher.  Whew!  Thank god it didn't cause a fire! For some unknown reason, the UPS blew up!

 

After I determined everything inside the house was safe, I went back to the closet, took the server out and tried to power it back on.  And of course, it wouldn't. Looks like the power supply and motherboard is fried. The jumpdrive with the unRAID software was toast too. I checked each data drive (6 WD 1TB Green drives), 3 were spinning up fine with all the data intact, and the other 3 (movie vault and parity) are dead!

AFIK, APC UPSes carry lifetime insurance against causing damage to your hardware and data.  Contact APC.

 

[prostuff1]

As purko suggested I suggest contacting APC about this.  There warranty should cover this sort of thing. If you really want to get anal you can probably stick them for the cost of the recovery of the data from the dead drives.

[FACP]

I thought about that.  Maybe I could get them to replace the UPS, Motherboard, PSU and hard drives.  As for the data, it was nothing that I couldn't replace.  Thank god it was the drives that had all my movies on.  Not the more important ones like family pics, important documents and a whole slew of files I need for work.  I could always re-rip my movie vault.  

 

I'll contact APC and see what I could get out from it.  But if it'll cause me so much of a hassle, i'll probably just let it go.  I'm just thankful it didn't burn my house down!

 

Thanks for the replies.

[Joe L.]

Your disks that did not spin up might just be fine if connected to a good power supply.  Do not count them out just yet.

 

It is very good to learn that you did not suffer damage to your home.  

 

No server is a replacement for backups, but it sounds as if you know that.  

 

You will need to get a new .key file from lime-technology.  email them and point them to this thread.

 

Definitely contact APC.  

 

Yes, you can just move the drives to a different PC/motherboard.  You do not have to re-load them.

Yes, they can be read individually.  

Right now, for you, I'd stay on the same release rather than complicate issues with an upgrade.

4.5.3 has a bug where disks appear as un-formatted when you first start the array.  You can simply "Stop" the array and "Start" it again to get the disks and the display in sync.  I've never seen it here on my server, but it appears to be an issue with users with some hardware.

 

A side effect of the bug is that the "Format" button is then made visible.  DO NOT PRESS IT, or it will format all those disks it is showing as un-formatted.  In reality, they just too too long to spin-up and get mounted.  This bug has hit several users when they were adding a new disk.  They thought they were formatting the new disk but instead formatted existing disks.  They were able to recover most(if not all) of their files, but is not not something you want right now to complicate your life.

 

As long as you know this, you can upgrade to 4.5.3.   Otherwise, use the earlier version.

 

Joe L.

[unraided]

Even if the hard disk itself is stuffed, their might be a possibility that just the actual logic board on the disk is just faulty if the disks are spinning (or not as it is controlled by the logic board anyways), but nothing is being detected on a new system with the old disks.

I've replaced the logic board on a faulty disk before and was able to salvage the data from it. You would have to buy another disk which is the same model as the one(s) affected, and replace the logic board (requires special torx screw driver set, which can be purchased easily). If the spindle motor isn't spinning after you replace the logic board, then you're in trouble I guess and you tried the cheapest way of trying to restore your data without seeking professional help which is very costly.

[purko]

I'll contact APC and see what I could get out from it.  But if it'll cause me so much of a hassle, i'll probably just let it go.

They are counting on it being a hassle, and on you letting it go because of that.   

 

[boof]

I have my unraid server in my walk-in closet plugged into a APC ES750 UPS.  When the power came back on, I heard a loud "pop" that came from the closet.  When I opened the door, I saw a little bit of smoke coming out of the UPS. 

 

Slight OT - has anyone else heard of APC's doing this? Always possible I guess but would be interested to know how rare it is. Slightly alarming!

[purko]

I have my unraid server in my walk-in closet plugged into a APC ES750 UPS.  When the power came back on, I heard a loud "pop" that came from the closet.  When I opened the door, I saw a little bit of smoke coming out of the UPS. 

 

Slight OT - has anyone else heard of APC's doing this? Always possible I guess but would be interested to know how rare it is. Slightly alarming!

 

I'd guess that has to be an extremely rare event, judging by the high dollar amounts they've listed in their Equipment Protection Policy.

http://www.apc.com/support/service/equipment_protection_policy.cfm

 

[terrastrife]

what psu did you have? if it was a power surge your psu shoudlve covered that with its own protection. unless it doesnt have it, that is

[Joe L.]

what psu did you have? if it was a power surge your psu shoudlve covered that with its own protection. unless it doesnt have it, that is

Almost any we would put in a unRAID server would have built in protection... many also are made to work on a wide range of voltages handling both 120 volts or 220 volt inputs without needing to be switched internally.    The surge from the power outage had to have been pretty serious for it to get past the UPS and take out the power supply.

 

I have a feeling the power surge from the blackout caused the damage and that the loud noise you heard was the metal-oxide-varistors failing from the over-voltage from the power company.    (They are designed to short when subjected to over-voltage, and then an associated fuse/circuit-breaker will open)  

 

You may find other electronic devices in the home damaged as well.  (Or other surge protectors that have failed saving the equipment connected to them)

[boof]

I'd guess that has to be an extremely rare event, judging by the high dollar amounts they've listed in their Equipment Protection Policy.

http://www.apc.com/support/service/equipment_protection_policy.cfm

 

I'm very cynical about those sort of things as the hoops you have to jump through and the things you would have to adhere to and somehow prove (or disprove!) mean there are lots of avenues for them to get out of a claim.

 

I'm sure it's not a commonplace event but the existence of some sort of policy like that doesn't reassure me it's extremely rare.

 

I'd much prefer to defer to peoples experiences. Though if it's the only failure of it's type anyone here has heard of than that's at least encouraging..

[yp_1]

I'll contact APC and see what I could get out from it.  But if it'll cause me so much of a hassle, i'll probably just let it go.

They are counting on it being a hassle, and on you letting it go because of that.   

 

 

Just read the fine print and you will understand why it is not worth it...

 

http://www.apc.com/support/service/equipment_protection_policy.cfm

 

 

Apparently you have to register your UPS within 10 days after purchase. How many people do that?

 

 

[prostuff1]

Just read the fine print and you will understand why it is not worth it...

 

http://www.apc.com/support/service/equipment_protection_policy.cfm

 

 

Apparently you have to register your UPS within 10 days after purchase. How many people do that?

 

 

 

I would imagine that not that many people read the little brochure that comes with it.  I have a tendency to read them, so i know that mine is covered should anything go wrong.

[boof]

I would imagine that not that many people read the little brochure that comes with it.  I have a tendency to read them, so i know that mine is covered should anything go wrong.

 

From my skim of the wording you're not covered for *anything* that can go wrong.

 

Only what APC deem as 'powerline transients'.

 

It would seem you have to provide them with the UPS and if they agree it has suffered as a result of a 'powerline transient' (and not because of one of the myriad of other exclusions they suggest...I have no idea how they would tell either way but I'm sure they can!) you then have to provide them with the failed equipment to check that that too has the effects of a 'powerline transient'.

 

This is of course all performed by them offsite and their definitions are also..defined..by them.

 

Again, I'm very cynical, but it would seem you have to essentially prove to them your physical installation has no issues and that what actually happened happened. How that can be proved by you by sending them your equipment I don't know - or to what degree of accuracy.

 

I *do* know that the document is full of ifs, buts and maybes. I don't believe they cover you for data loss or incidental damage - which means if the UPS goes pop and burns your house or premises down...?

 

It reeks of marketing to me rather than any sort of firm guarantee (far too heavily balanced towards them than the consumer) and I'd be very interested to hear of the number of claims versus reimbursements that have ever been made - should such numbers exist openly.

 

I hope I'm wrong as an APC is top of my list of things to buy - probably this month - due to some interesting breaker issues in my new house.

[bubbaQ]

I second the notion that a drive may appear dead, but will be OK if you put it in another system on another PSU and controller.  I have seen that many times.

[terrastrife]

i must be the only oen reading 'hdds have been tested on another psu already'

 

maybe im wrong

[Bizarroterl]

FACP - please post APC's response once you have it.  I'm using the exact same UPS and want to be sure they'll back up their claims. 

 

BTW - I opened a support case with APC.  I needed a rs232 cable to hook up my 750 to my server (ESXi Host/Win2008 client) so I asked for a PN or a wiring diagram.  They're sending me a cable at no cost, and my 750 is at least 6 months old. 

[bubbaQ]

i must be the only oen reading 'hdds have been tested on another psu already'

 

No, I saw it and my advice still stands... try a different PSU and controller.  A latch (short) on 1 pin can prevent a spinup.

[DoeBoye]

Slightly off topic, but I thought I'd throw in a word of encouragement, as far as contacting big faceless corporations about promises they make...

 

Here's my story:

 

A few months ago, my plasma tv failed and of course, it was out of warranty. Having been bombarded over the last few years by the credit card companies and *their* markety-type extended warranty advertising, I thought I'd give it a whirl...

 

I was, to say the least, extremely pleased by the process. In the end, I spent maybe 40 minutes on the phone (including on-hold times), and a couple of emails to fix a mistake I made in sending them the required paperwork, and they approved the coverage.

 

3 weeks later, a check for $600 arrived at the door to cover the repair.

 

I guess the point I'm trying to make, is it might be worth at least a call. The fine print may be to cover their asses in case of enormous claims or scammers. For relatively small claims like yours, they may be much more lenient...