Tag Archives: hardware

Drive to live?

hddOne of the very valuable things to come out of large data centres is large-scale reliability statistics. I’ve written before about my suspicions that my Seagate drives weren’t as reliable as they might be, but I had insufficient data for this to be anything other than anecdotal.

And then a couple of weeks ago, I pulled a couple of old 2.5″ drives off a shelf — Western Digital ones, I think — intending to reuse them for backups. They both span up, but neither would work beyond that.

So I was very interested by this Backblaze blog post which discusses their experience with a few thousand more drives than I have at my disposal. They use consumer-grade drives, and are very price-sensitive.

A quick summary:

blog-survival-drives-by-month

Some quotes:

Hitachi does really well. There is an initial die-off of Western Digital drives, and then they are nice and stable. The Seagate drives start strong, but die off at a consistently higher rate, with a burst of deaths near the 20-month mark.

Having said that, you’ll notice that even after 3 years, by far most of the drives are still operating.

Yes, but notice, too, that if you have four computers with Seagate drives, you should not expect the data on one of them to be there in three years’ time. And, quite possibly, not there by the Christmas after next.

The drives that just don’t work in our environment are Western Digital Green 3TB drives and Seagate LP (low power) 2TB drives. Both of these drives start accumulating errors as soon as they are put into production. We think this is related to vibration.

and

The good pricing on Seagate drives along with the consistent, but not great, performance is why we have a lot of them.

If the price were right, we would be buying nothing but Hitachi drives. They have been rock solid, and have had a remarkably low failure rate.

and

We are focusing on 4TB drives for new pods. For these, our current favorite is the Seagate Desktop HDD.15 (ST4000DM000). We’ll have to keep an eye on them, though.

Excellent stuff, and worth reading in more detail, especially if longevity is important to you. It’s tempting to fill old drives with data and put them on the shelf as archival backups, but this would suggest that you should only use new drives for that!

Oh, and if you’re wondering about which SSDs to buy, this report suggests that Intel ones are pretty good.

Update: Thanks to Dominic Plunkett for the Backblaze link, and for Rip Sohan for a link in the comments to the TweakTown article that attempts (with some, but not a great deal, of success) to debunk some of this. The previous article I mentioned above links to an older Google study which didn’t distinguish between manufacturers and models, but did say that there was a correlation between them and the failure rates. It also catalogued failure rates not too dissimilar to the Backblaze ones after 3 or so years, so the general implication for home archiving remains!

A most sinister thread…

One of the things we try to do here at Status-Q Labs is to reduce the amount of frustration experienced by our fellow men and women in their daily lives.

Take, for example, the case of reverse threads. This is not, as you may suppose, an advanced mode of needlework, but rather the practice of using screw threads which turn clockwise to undo or loosen, and anti-clockwise to tighten, something that observant readers will detect as being contrary to the natural order of things. There are good reasons for using these — also known as left-hand threads — in situations where the normal use of the the device would tend to cause it to unscrew of its own accord.

However, there are few things more frustrating than not knowing that the thing you're trying to unscrew is, in fact, tightening up, especially if it's old or rusted or damaged and you expect it to be somewhat tricky anyway. Here are a few situations where I've encountered reverse threads in normal life: remembering these may improve the level of your future happiness and avoid some skinned knuckles and unwarranted expletives.

  • Left-hand bicycle pedals. These are screwed into the crank with a reverse thread. Which is a little strange, if you think about it, because their rotation on their spindle is clockwise, but there are other effects at work. More information here.
  • Gas cylinder valves. This is for safety, rather than mechanical reasons. Combustible gases, such as the propane or butane you might connect to your caravan or gas barbecue, use left-handed threads, so you can't accidentally connect them to things expecting an inert gas. This is a good idea, but I can't help wondering how often frustration has caused people to start hitting spanners attached to explosive cylinders with heavy objects…
  • Drill chucks. I had to replace the chuck on my hammer drill recently because the jaws had seized up. It screws onto the main spindle of the drill with a standard thread, but is held in place there by a bolt which goes the opposite way. (I guess you only need this on a reversible drill!) Incidentally, even when you know which way the threads go, chucks tend to be fairly firmly fixed, and I might not have managed it if my friendly local hardware store hadn't shown me the allen key trick.

So there you go. Lodge those in your little grey cells and one day, I promise you, you'll thank me!

Anyone know any other situations where the unsuspecting might encounter left-handed threads in normal life?

Update: Lyndsay Williams pointed me at this Wikipedia page, which lists under 'Handedness' a few other places where left-handed threads are used. My favourite is that lightbulbs on the NYC Subway used to have reverse threads, so you couldn't steal them and use them anywhere else!

 

 

Love and marriage, love and marriage…

…go together like a RaspberryPi and Veroboard…

“The thing people don’t understand about weddings”, said a perceptive friend once, “is that they think it’s ‘the bride’s special day’. When in fact, of course, it’s usually the bride’s mother’s special day. It’s when she gets to create the wedding for her daughter that she wishes she’d had herself. And she’ll be able to remember the details of this one.”

The male equivalent is probably buying a model railway set “for the benefit of your children”. Or, at least, it used to be. Now, of course, geeks of my generation are terribly keen to support the RaspberryPi, “because of all its educational benefits”.

I was thinking about that this morning as I soldered transistors onto Veroboard… for the first time in about 30 years. It’s for the educational benefit of my dog…

Using Little Computers to control Big Computers

Here’s my latest Raspberry Pi-based experiment: the CloudSwitch.

I don’t discuss the software in the video, but the fun thing is that the Pi isn’t dependent on some intermediate server – it’s using the boto module for Python to manage the AWS resources directly.

I decided to build the app slightly differently from the way I would normally approach a little project like this. I knew that, even for this very simple system, I would have several inputs and outputs of various kinds, some of them with big delays, and I wanted to make sure that timing hiccups or race conditions didn’t ever leave the lights displaying something that didn’t represent reality.

So this is only a single python file, but it runs several threads – one that looks for button presses, one that monitors and controls the Amazon server, and one that handles the lights – including flashing them in various patterns. They interact with the main thread using ZeroMQ messages, which is a lovely way to do inter-thread communications without all that nasty messing about with semaphores and mutexes.

Update: Here’s the very simple circuit diagram. The illuminated buttons I used have LEDs which take a little more power than the Raspberry Pi can really drive, so I put a couple of NPN transistors in there. It really doesn’t matter too much what they are – I used the 2N3904.

Disk Risk

Mmm. I seem to have had a lot of hard drive failures recently – Seagate drives, mostly, though, to be fair, the majority of my drives are Seagate just because my favourite supplier happens to like them, so I would expect see more failures there. The last one, though, is just 18 months old and has started making ominous clicking noises. They don’t make ’em like they used to. Stuff I’ve read online tends to suggest that it’s hard to assign blame to particular drive manufacturers, but particular models do tend to have rather different failure rates.

I do, I realise, have rather a lot of hard disks. I have three 4-bay Drobo enclosures, for a start, so that’s 12 drives even before I start adding on the miscellaneous backup disks, TV-recording disks, etc. Not to mention the internal ones in all our various machines. There must be 20-25 hard disks around here, and even though manufacturers’ specs talk about a <1% annual failure rate, studies tend to suggest that real-world figures are rather higher. One of the biggest studies, done by Google a few years ago, showed failure rates of 1.7% in the first year, rising to over 8% in the third year.

Yes, many of my drives are about that age, so if I really have 25 of them, I guess I should expect one to die every six months or so. Bother.

This suggests to me that money spent on things like my Drobo enclosures is worthwhile, because, though they are pricey, especially once you’ve filled them up with drives, any single drive failure is unlikely to be catastrophic – as disks die, you just replace them with whatever size is currently in vogue. My main Drobo currently has two 2TB drives, one 1.5TB, and a 1TB. There are those, I know, who have had less positive experiences with some Drobo kit – I found a DroboShare networking add-on to be decidedly wobbly at a past company – but in the simple use case of a Drobo plugged into a computer, I’ve been very happy and have replaced several drives without ever losing data.

The other thing that the Google study found was a strong correlation between when disks start reporting errors (which they can do using the S.M.A.R.T technology built into modern drives) and a failure soon afterwards. It’s worth, therefore, having something that checks the S.M.A.R.T status and lets you know about issues as soon as they are reported, even if the drive is still apparently working OK. On the Mac, Disk Utility can tell you about issues, but only when you go and look, so I use SMARTreporter to give more regular checks.

OK, things are getting better. There is another issue, though.

On the Mac, at least, most external drives are connected by USB or Firewire, and in general S.M.A.R.T information is not read through those interfaces – if you look in Disk Utility, you’ll see it’s ‘Unavailable’. More sophisticated enclosures like the Drobo will check the S.M.A.R.T status themselves and warn you when things look dubious, but your average USB-connected backup drive may give you no such warnings.

So I was interested to discover this kernel driver project which enhances the standard OSX USB and FireWire drivers to make S.M.A.R.T available for a lot more interfaces. (Download v0.5 here). I’ll try it on my Media Mac Mini, which has three external drives, and see how it goes…

© Copyright Quentin Stafford-Fraser