So, I’ve had  a fun expensive time of things lately.  Some of it has just been the luck of the draw when dealing with a house that is 10 years old.  No, it’s not an old house, but that is long enough for things to start breaking. 

Originally published at Angrywaffles. You can comment here or there.

md0 : active raid5 sdd1[1] sdi1[0] sde1[5] sdc1[4] sdg1[3] sdh1[2]
      2441919680 blocks level 5, 64k chunk, algorithm 2 [6/6] [UUUUUU]

Originally published at Angrywaffles. You can comment here or there.

Woke up this morning, and my fileserver had emailed me this:

md0 : active raid5 sdh1[6](F) sdd1[5] sdc1[4] sdf1[3] sdg1[2] sdi1[7](F)
      2441919680 blocks level 5, 64k chunk, algorithm 2 [6/4] [__UUUU]

Looks like I get a trial-by-fire of MD RAID recovery tonight, once I figure out what broke.

Originally published at Angrywaffles. You can comment here or there.

Not bad for an affordable motherboard (Gigabyte EP45-DS3L, $85 or so after rebate), some cheap Kingston DDR2-800 RAM, and a $30 heatsink.  Core voltage is only 1.28V - I’m not going to bother pushing it harder because at this point the RAM is already overclocked and I can’t run the RAM any slower than the current speed.  I’m happy with the speed, a 3.1GHz quad-core Core2 is quite the upgrade from my 2.5GHz dual-core Opteron.

One of these days I’ll probably slap a new video card in here (current one is getting a bit old, 7900GS) but I’m in no rush.  Probably when the next-gen midrange gets cheap.

Originally published at Angrywaffles. You can comment here or there.

So, I’m currently running two fileservers.  One at my house which has a 4×500GB RAID5, a 2×250GB RAID0, and runs a few other things for the local network.  Pretty much anything of any relevance gets stuck on the server, the actual computers just get by with whatever minimal hard drive was cheapest.  The other I stash at my mom’s for backup purposes, just in case something truly nasty happens that physically takes out my box at home.  It was originally set up with a 4×200GB RAID5.

Originally published at Angrywaffles. You can comment here or there.

Originally published at Angrywaffles. You can comment here or there.

Thank goodness for eBay.  I’m still rocking Socket 939-era hardware on both my desktop and Kate’s desktop, which unfortunately means that new motherboards are simply not available at reasonable prices anymore - so my only option for a replacement motherboard was eBay.  Under $30 shipped for a decent AsRock board, the 939SLI32 or whatever they call it, it’s some strange ‘bridge’ board which can have an optional daughtercard installed to use a Socket AM2 / DDR2 setup instead of Socket 939 / DDR.  It’s also ULi’s last chipset prior to being bought out by nVidia, and it apparently features some ‘unofficial’ SLi implementation that’s probably not very well supported these days.

Amusingly enough, though, it’s a solid overclocker, even moreso than the board it replaced (DFI Lanparty NF4 Ultra-D) was.  With no other changes I now have the CPU running on a 280MHz FSB - 2.5GHz clock, putting it roughly on par with an Athlon64 5000+.  It would only do 270MHz on the DFI. Not bad for an old server chip.

The old motherboard was definitely the problem, but by no fault of its own.  I had to replace the chipset fan once a few years ago because the factory one got noisy - apparently I didn’t make sure the pins were still secure since one failed and the heatsink lifted clear of the chipset with no warning.  It’s toast.

Now, if I could only figure out why Folding@Home stopped running on my webserver for no good reason…

Originally published at Angrywaffles. You can comment here or there.

Today, not long after getting home, the power blinked - just long enough to shut down the TV and a couple of computers, including the media center (which has a crappy power supply so that’s no surprise) and my desktop (which was a surprise).  Every desktop system I have aside from the HTPC uses relatively high-end power supplies, which I’ve noticed have the side benefit of keeping the system on a bit longer in the event of a short gap in AC power.

After quite a few futile attempts to get my desktop to boot again, I’ve concluded it has now developed a nasty hardware failure somewhere.  The random nature (it hangs anywhere from before POST completes to after logging into Windows) makes me suspect the motherboard, but I suppose the power supply could be at fault as well.

It’s really a pain because I really don’t have much time to sit down and diagnose why it’s broken, but at the same time I don’t have a ton of money laying around to throw at problems.  I’m actually somewhat tempted to just let the damn thing sit until I can afford to do the upgrade I was planning on - a Nehalem-based quad-core Intel setup.

Originally published at Angrywaffles. You can comment here or there.

Originally published at Angrywaffles. You can comment here or there.

The hard drive industry has been a colossally screwed up place for ages. It’s better now than it was in the past - but it’s not exactly good now either. At least, I suppose, we aren’t living in the dark ages of strictly one-year warranties combined with sky-high IBM 60GXP and 75GXP Deskstar failure rates.

Not that the primary hard drive in my desktop seems to have gotten the notice. I spent a few hours today reformatting it and installing a fresh copy of the OS, which seems to have done in the old drive - a Seagate IDE 200GB drive, which is now chucking up SMART errors on boot and every so often while in Vista. I doubt it will give me actual data corruption in the immediate future but if a SMART parameter is far enough for Vista to be complaining about it, it can’t be good. I at least don’t keep any vital data on it - that’s all on the fileserver on a RAID5 array, with the really critical stuff backed up to a separate off-site server on its own RAID5 array. But I digress, and I should probably post about that later.

So, the drive is dying, (though not completely gone - I’m typing this up on the ‘dying’ drive!) so I pull up the serial number and go to Seagate to check the warranty status. Their warranty and the support I’ve gotten in the past is part of why I bought this drive…back when a lot of companies were cutting to one year warranties, Seagate went to five-year warranties. I check and yes, this drive is still under warranty until sometime in 2010. I go to start the RMA process and, as I’ve always done in the past with Seagate, Maxtor, and Western Digital , I go to set it up as an Advance RMA - I get the new drive before I ship the old one back, and I also get manufacturer-approved packing materials to ship the old one back in. Really, a no brainer.

Anyway, I go to check that box and - they want $19.99 for this privilege! Plus they’re plugging a ridiculously priced upgrade (wow, I can send you a 200GB drive AND $50 and get a 250GB drive back in return?) and a $15 fee to have your shipping upgraded from slow to less-slow.

The whole thing has soured me on the one reason I have been buying Seagates - the warranty. I’ll be picking up a drive by other means (probably a used one, I have no need for a new 500GB drive and you don’t get to proper cheap $/GB ratios until that point these days), and when I get the RMA back on the 200GB, it’ll go into that offsite backup. No more going out of my way to buy Seagate, and I’ll probably make a point to buy other brands when given the choice.