The Fault in Our Hardware

On Sunday afternoon I knew I had a dead power supply.  So I replaced that.

That evening I was pretty sure I had a toasted motherboard as well.

Doing research on Monday I figured out that my old Intel Core i7-950 required a socket no longer readily available.  So I headed on down to Fry’s for a new motherboard and processor.  I ended up with a Gigabyte GA-Z97X-UD5H and an Intel Core i5-4590 3.3GHz Processor, both of which were on sale at the moment.  Not sure if that i5 was a step down from the older i7.  My hope was that it was at least on par with four years of updates.

It Also Gets HBO

All of this is junk now…

Then I spent Monday afternoon and evening putting everything back together, only to be stuck for an hour trying to get the system to boot.  I was worried how Windows 7 might behave with new hardware under it, so I got out the install DVD and tried to update it or, if I had to, install a fresh copy of the OS.  Only after tinkering with that did I discover my next problem.

My system had two 1TB drives in a RAID 1 configuration, which is essentially two drives mirroring the same data, so if a drive goes bad you still have everything.  I mean, what are the odds of both fucking drives dying at once?

The odds are greater than zero it seems, as I found that neither drive would spin.

This is where my heart begins to sink and I start feeling sick to my stomach.  Up until this point it was all just a technical issue.

I have external drives hooked up and key data backed up elsewhere.   All of our digital pictures from the last 15 years are copied to multiple drives.  But a lot of my files were not backed up.  I had not, for example, moved screen shots or notes in Notepad ++ or any number of other things off the main drives for a while.  All of that is gone.  Lots of data, lots of images, lots of notes and knowledge.  I had full installs of Warhammer Online, City of Heroes, and Star Wars Galaxies.  Not that those are all that useful, but I won’t be getting those back.

And don’t get me started about my iTunes library.

That means everything had to start from scratch again.  There will, no doubt, be a blog post about what games I choose to install again and which games I let fall by the wayside.

I went out again and bought a 240GB SSD and a 1TB drive.  No more redundancy, but at least a faster boot time.

I got the operating system installed, then the drivers for the various devices on the motherboard, and then started loading the updates.  So many updates.

I got browsers install, but with the data loss all my bookmarks and whatever stored passwords I had are gone, so it is pretty much up to my memory at this point.  For example, I bought Microsoft Office online and should be able to install it again… only I cannot remember the account I created for that.  That is in a saved email somewhere, gone.

I am happy now that Trillian, which I use for IM, saves your settings at their site.  I was able to get up on IM quickly enough.

Eventually I got to the point where I thought I could install a couple of games.  I chose the two on my active list, EVE Online and World of Warcraft.  WoW has a new expansion running and our guild is back and online a lot, including Gaff, who returned to WoW, while there are two CFC deployments going on in EVE and constant pings for fleets.

However, I noticed a new problem.  The system, which comes up fine and appears to run well, restarts without warning every so often.  This happens rather quickly if I run WoW, I cannot, as a measure of time, harvest garrison resources before the system begins to chunk, giving me less than a minute’s warning that it is going down.   But this restart also happens with EVE or if I just leave the system sitting idle, it just takes longer.

At this point I have replaced everything except the video card, the RAM, and the case itself.  I suspect there is an issue with the RAM.  When I first loaded up the motherboard it indicated a fault initializing until I swapped around a couple of the sticks.  So it is off to Fry’s again at some point today to buy another pair of DIMMs.  But if that doesn’t solve the problem, then I am hosed.  I don’t know what I will do then.  I will have blown my discretionary budget on a bunch of parts that do not add up to a whole.

17 thoughts on “The Fault in Our Hardware

  1. Rohan

    Well, it’s probably too late to help you, but my philosophy is that if the power supply or motherboard dies, the entire computer is pretty much a write-off. There’s too much chance of cascading damage.

    Consequently, I’ve tried to better than average for those two components when buying a computer.


  2. evehermit

    The energy spike when a power supply dies can certainly damage components in your PC – I have experienced what you have, motherboard, memory and video card that become flaky and unreliable. You end up having to replace everything.

    (I have a friend who replaces his power supply every two years as a precaution.)

    I also never use RAID in my home PC’s. The chance of full data loss is greatly increased over just using a standard disk configuration. At least with a standard disk you can plug them into an external enclosure or cradle or use free data recovery software when something goes wrong with your PC. With RAID disks you are often stuck trying to find the exact same hardware controllers and drivers, and if they are corrupt you often can’t get anything off them. (Been there and done that many times over, mostly through work.)

    Similarly for backups I use (in part or whole) end of day file synchronisation backups of email and data. Being able to see the backup files without having to install backup software and run restores greatly increases the chance of accessing your data after an issue.

    At least you should love the speed increase with the SSD. Just make sure you have your backup regime in place quickly. I’ve had 2 of the 5 SDD’s I’ve used die on me.


  3. p0tsh0t

    Sounds suspiciously like a cascading failure. I’m not buying two hdd cratering at the same time without a common cause. If they’re not spinning, the good news is that the data is likely undamaged and its probably a diode(s) issue that blew due to overvoltage.

    I’m betting most will be recoverable albeit with some voodoo. Dig out that multimeter!


  4. Wilhelm Arcturus Post author

    Actually, my next step is to pull the video card, since it was yet another device plugged directly into the power supply. The motherboard/chipset I bought has built-in video which, while not great, is enough to drive WoW. And it even has DVI output. I’ll try running WoW for a bit with just that. If the video card is the problem, then I can always dig out my old GTS 450 and install it again.

    The hard drives… yeah, I suspect that the data is fine, but getting the motors spinning again without paying somebody a grand… which seems to be the going rate for that sort of thing… is the trick.

    @evehermit – Actually, RAID1 is just mirroring with multiple drives. Broken mirror operation is pretty reliable. In fact, one group at my last company used to use broken mirror as their roll-back strategy for upgrades… because they were a bunch of cheap skates. Pull one drive from each pair out of the SAN, do the upgrade, if it fails, restore from the drives pulled offline. So if just one of the damn drives was still running, I’d be golden. Such is life.


  5. whorhay

    There are diagnostic tools you can use to test your RAM, though they might not be able to complete before your system crashes. But I would expect that they could probably tell you if the RAM is the problem pretty fast if they are what’s crashing you.

    I use a SSD which is mirror’d to a partition on a conventional HDD.


  6. Kruppe

    wow, good luck. Losing data is my biggest fear. I am getting ready to build a new computer and hand off my older one to my daughter. I am trying to figure out how to move all my stuff now.


  7. ytypo123

    Sounds like ram to me. Same thing happened to me on my computer. Ram was bad. Computer would randomly reboot no bsod or anything.


  8. zaphod6502

    Lost too many drives over the years to no longer trust them. I’ve tried raid configurations and that was even worse. Nowadays I back up all my important data and photos to DropBox (I am on their 1TB plan) and Photobucket.

    As for upgrading piecemeal I feel your pain. I tried this after a motherboard failure 7 years ago and I ended up replacing all of my core components and wasted a whole lot of money trying to diagnose a dying computer system.

    Now I just upgrade my computer system every three years (mobo+ram+CPU+GPU) regardless of whether it needs it. Saves me waiting for it to die and I always keep up to date. I sell my old parts into the used market to offset the purchase.


  9. Isey

    I use Onedrive (1TB of storage when you own Office on the annual plan, which I do for work) and it lives both in the cloud and live on your drive. So things like pictures, etc. never disappear. It’s been really handy for personal files.

    Gaming files and the like are easily reinstalled. Sorry to hear it, what a pain!


  10. thighzen

    I’ve been in your situation a few times as well. I don’t have the knowledge or tools to nail down the problem for certain so I end up troubleshooting in a similar fashion to yourself. Can get quite frustrating when the first few parts you buy don’t fix the issue.


  11. Tolga

    Instead of raid, I’d suggest a backup system or plan. Works ultimately better for home/small office users.

    And for a quick quality of life hack; use chrome for browser. it remembers all your bookmarks (and passwords if you set it). Even if you don’t like chrome, installing and exporting your bookmarks into it once in a while ensures you to have bookmarks available to you in an emergency.


  12. Samuel Silbory

    i5-4590 is a much better cpu than the i7-950. Same number of cores, more efficient, faster in pure GHz as well…. The only thing it doesn’t have is hyperthreading which on a 4 core cpu for desktop use is near useless, and the i5-4590 would still outperform in the corner cases where hyperthreading mattered. Also I’d try the builtin video as it’s certainly enough for Eve, and Wow.


  13. Atruin

    On the hard drives, did you try to connect them to power supply without the SATA cables connected and do they spin up then (you can feel the gyro effects of the disks inside even if you don’t hear anything)? I find it very unlikely that a dying mainboard power supply would take them out both at the same time. It might be an issue with the RAID controller of your new board. It sees there is a RAID on the disks and refuses to power them up, because it uses a different way of calculating checksums and stuff. I have seen that on our company RAIDs as well. A solution might be to get a second hand working motherboard that uses exactly the same SATA/RAID controller like the one you used before.


  14. Solf

    Hopefully by now it is of no relevance whatsoever, but just in case / for the future…

    A failure that happens ‘pretty quickly if you run ‘heavy’ app (WoW)’ may indicate problem with CPU cooling. These days it is easily checked with likes of e.g. SpeedFan.

    Coincidentally — is ‘subscribe to comments’ option is gone (again)?


Comments are closed.