Apple's Time Capsule, which went on sale last week, may not be quite what the Mac maker maintains it is, in the storage department at least. Time Capsule is Apple's 802.11n Wi-Fi router with an integrated 500GB or 1TB "server-grade hard disk drive", to use Apple's words. But when one buyer took his TimeCapsule apart, he found …
Mean time to failure comparable to server-grade HDD
Apple have said that the mean time to failure of this hard drive is comparable to that of server-grade hard disk drives, giving their basis for calling it 'server grade'.
possibly the least 'server grade' drive
I hope that IBM have moved on from the 75GXP 'Deathstars' made (in)famous by the click of death....
A backup solution with one drive is ridiculous anyway.
I see Apple are keeping up with their commitment to 'quality' products. They could only have done worse if they put in a fujitsu drive or a western digital. Many many peeps are suddenly going to lose their backup with no warning, just a sudden loud clunk clunk clunk noise coming from their box.
Its the same drive that ships with the Xserve.
It seems a cheap way around it, sure, but they never actually specify what server grade means anywhere in the blurb about the drive.
Paris, because I'm sure if she bought a Time Capsule she'd be concerned and confused over the meaning of the term server grade.
Anyone have any Mac servers?
What do they put in them? It they're Hitachi too then you can't fault the wording.
Of all the brands they could've picked though, as already said the Deathstars are infamous.
What's a server?
The very first server I ever built was just a desktop machine, but since it provided central file storage and printer sharing, as well as simple redundancy and backups, it was a server.
It also used desktop drives.
On the other hand, I agree with Stephen - if you're waving around 'server grade' disks, then you shouldn't be using one drive anyway. It's a bit like building a high-end gaming PC with "only" 256Mb of RAM: it's a half-finished job.
Depends on the server
These are the same drives in the Xserve. Therefore, by a retroactive marketing definition, they're server-grade.
No, these are not datacentre hard drives, not by a long shot, but they're better than are in most home-user's machines. We're getting to the point where we need to draw the line between home servers and enterprise servers.
A server is a machine that serves data and this drive is ideal for that, but it's not what people think of when they think of enterprise-level servers. However, I would imagine that this drive is better quality than most of the drives in servers out there today. It's the kind of hardware you'd expect in a device of this price, aimed at the market it's aimed at.
MTBF and reliability etc
Just a guess here, but I think I can probably explain it. The difference between the "server grade" and desktop Hitachi drive is probably minimal, as pointed out by the article. The charge on top of the server grade is most likely down to the warranty costs. Does anyone know if there is a different MTBF for each drive for example?
To me, server grade wouldn't imply it was more reliable, merely that it was faster for r/w access and seek times. In a server environment, you wouldn't be relying on a single disk in the first place. Any disk I've ever used (server or desktop) either fails in the first few months, or not at all for the accepted life of the product. Rule of thumb for desktops for example, if it lasts the first 3 months, it'll probably make 3+ years.
Looking at Hitachi's description of what it the drive should be used for (network storage servers), I think Apple are quite within their rights to quote it as such in this case.
Is this drive really likely to spin 24/7? Not very 'green' if it is.
backups with 1 disk are way cool
errrr Any backup has at least 2 disks - the copy and the source; what about sites - is a backup really a backup if the source and copy are in the same room?
Single drive backup can work
For home use, a single drive backup can work. Assuming users only use it as a backup drive (ie there is no original data on the disk) then all it can do is be good. If your main machines hard drive fails, you have a backup. If your backup drive fails, you still have your main machine.
It isn't redundant and it isn't perfect. For the average user, the time between a failed drive and a replacement (and having a backup again) will probably be weeks or months, rather than the hours that you'd get in a server system (which would no doubt have more than 2 disks for backup and live purposes.
Ideally, you'd be running a RAID 1 system on your mac, have a backup solution with some form of redundant RAID and also have a nightly offsite backup, but for home use, is that likely? Surely having some sort of backup (even time machine) is better than having nothing at all.
here we go again!!
ready for the Apple bashing!!!
@Stephen, I guess you dont use a laptop as a mobile device?? I do, its out everyday, and quite heavily abused... when i get home, I dont have the time to remap the drives and run a backup to my server... (which I used to do)
but I do use the laptop in the evenings... the idea of my laptop just automatically backing up as i get in and turn it on, all transparently is ideal!
yes, the one disk, onsite backup is u/s if the house burns down with the device and client inside... or both disks die at the same time...
but the more common data loss would be me deleting the wrong file/folder (easily done), dropping my laptop (this has happened) or someone stealing it... In that case the single disk, always on backup will be so much better than my raid array, that I forgot to connect to last night because I was running late and needed to catch Scrubs!!! :)
the issue here is all to do with the claims of TYPE OF DRIVE!!! and this is a nice grey area :) specs read server grade, label says desktop!!!
not that id ever dream of putting less han a 10k drive in a server! :) but realistcally seek times arnt really an issue in this application....
7200 RPM - Server grade
Crikes, I know Apple make some claims, but we haven't used 7.2's for years, 15k's now baby !
It Works for Its Target Market
I think consumers where expecting something along the lines of Seagate's ES.2 drives. Basically, higher grade 7200rpm SATA drives with a much improved MTBF. I doubt anyone was expecting it to have SAS drives in a RAID configuration. It's essentially a low-cost NAS that incidentally provides a convenient place to store backup files.
Yes, they could have made it more redundant. Though, it would really only be problematic if your Mac took a crap at the same time your Time Capsule did, but the chances of that are slim enough that it's not really an issue for a consumer-targeted device.
Personally, I fall into the 0.1% of users that could care less if their machines takes a crap (aside from the hassle of getting it running again). There's nothing on it's HDD that I would lose any sleep if it was lost. I'm not a music fanatic with gigs of MP3s, I keep my digital photos on an online gallery site and not my local HDD, and I prefer my movies on DVD. So, there's nothing there worry about losing anyway.
A good thing to think to yourself when determining what sort of backup solution you need is, "If my PC/Mac took a crap tomorrow, how f**ked would I be?"
MTBF = Weasel words
Let the Weasel Words commence! Apple fanboys and spinners alike will now try to avoid the obvious. The obvious? Apple + Time Capsule + Server grade drive = Busted!
If it's server grade that would mean that someone, presumably Apple themselves, would use the drive in a server. If no one does then it's not server grade, merely quite reliable. (Note : "quite" != completely)..Don't know about you, but 'quite reliable' isn't a description that would make me buy a backup box. Server grade might though. Well, it would if I wasn't a cynical geek that would build his own backup box with mirrored drives.
server drives and server grade drives
server drives spin at 10K RPM + ... because of throughput speed and access time (seek time). It's hard ot find these beasties larger than 130 GB ... they invariably use 1.8 , 2 inch or 2.5 inch platters.
Server grade drives are drives that can be used for non-Throughput critical applications, such as archiving and long term storage. So apple can indeed claim this is a 'server grade' drive. Hitachi markets this thing as 'server grade' after all..
Apple has ALWAYS Cheated on Quality Components!
Unless you're a total Apple Kool Aid Drinker Droid, even Mac buyers over the last two decades KNOW that Apple has ALWAYS used substandard HDD's (size and brands), too small and slow RAM, and WAY behind the Tech curve CD/DVD drives. And then the greedy bastards at Apple over charge by 2x for upgrades like RAM ,etc.
Get the point? Apple = greedy bastards and NOT a leading edge tech purveyor.
Ps. Time machine - ANOTHER Apple iNOvation despite all the clamor, is nothing more that a NAS with BU and step-back software which has been around for Windows for over a decade.
Maybe people have just started using lousy drive in servers...
merely playing warranty games
I doubt if this label means it will fail more or less often - just that you'll send it back to a different place (Apple) if it breaks.
In practice anyone who runs a server without some form of redundant storage is, IMHO, a fool anyway. So I reckon just having a single disk in a box, makes it a domestic appliance - not a commercial grade server.
I think this is pretty crap but let's not go nuts: it's not "a backup solution with one drive". Since it's external there's two drives involved and the chances of both yourinternal and the external going down at the same time are remote for most Mac users who don't tend to do really heavy work anyway.
And in other news..
Apple computer announced today the invention of the server grade iLED. The latest must have gadget will allow users around the world to see server status without having to log in to their systems. "There have been other attempts to use visual aids for server grade reporting," Jobs was quoted, "however, our superior 'iLED' not only comes in a sleek minimalist white package, but it outpaces all other attempts at providing instant, server grade status of systems." Apple has not released pricing to date.
Small issue, but still an issue
The problem isn't the quality of the drive, or even how it can be argued (as a fudge) to be 'server grade'.
The issue is that Apple's advertising for the product is intentionally misleading to readers. It is short and simply a direct equivalent of Microsoft and 'Vista Ready'.
The spec Microsoft set for 'Vista Ready' was high enough that Vista could indeed run, how ever it certainly didn't run in a way that consumers expected 'Vista Ready' to mean.
The short and simple: Apple is misleading buyers, which is a shame (especially as I actually think the device is a good device regardless).
Ahhh, the trusty old deathstar
I'm pretty sure apple have just gone to an office liquidation sale and bagged all the old desktops, i'm pretty sure every company owned machine since 1995 had one of these, or a quatum at the very least.
Anyway, i've had two deathstars over my life time, and both at some point decided to self destruct (in different machines at ver diffrent times) server grade my puckered sphincer!
The 500GB ...
The 500GB Time Capsule uses a ST3500630NS:
"The Seagate® Barracuda® ES enterprise hard drive is the industry’s most reliable, highest capacity 7200-RPM hard drive for 24 x 7, multidrive, business-critical applications. The Barracuda ES hard drive is built with perpendicular recording technology for maximum capacity, reliability and cost-effectiveness."
Hitachis not so bad now
I too agree when it was IBM the Deathstars were shocking, I had 2 of those bad boys fail on me. However I have been running several of them now they are Hitachi (touch wood) they have stood up to some serious hammering with near 24/7 downloading - you know with the ocassional power cycle while the windoze updates go on ;o)
They said the disk was 'server grade' and it is, as in: 'used in servers', the specs and the manufacturer also tell the same story.
I'll admit there's specialist areas where 15K disks make all the difference but for the vast majority of servers on this planet they're overkill, on this box though, which isn't a server and doesn't claim to be*, they'd be a retarded way to double the price, quarter the capacity, and demolish the sales.
* get a fecking grip, it's a consumer network backup box FFS.
What did people expect?
Its just a networked HDD with the usual Apple ease of use + mark-up.
Who keeps important data in a single location anyway? I bet that the data stored on El Reg's 'server quality' server HDDs are backed up onto tape! No HDD is THAT reliable.
Having said that, I'd be happier to see a cool and quiet Samsung HDD - especially if this was in my front room.
I think the "8 hours a day" rating was invented for the Deskstar 75GXP and 120GXP (aka Deathstar) back when IBM couldn't keep them working longer than a year.
What does Hitachi say?
It's amazing isn't it? Just mention Apple and.......
anyway, here's the Hitachi spec.
Which means what?
And "Server Grade" means what, exactly?
Wait a minute...
...not a single post from Webster Phreaky? It's getting so you can't count on anything anymore.
yawn . . .
1. The leading reason why home and small business users lose data is because most haven't got a backup, or when they claim they have, it turns out not to be usable. So credit to Apple for actually producing such a convenient and usable backup solution.
2. Hitachi provide identical drives for desktop and server. (Hitachi Deskstar 7K1000 / Ultrastar A7K1000). The difference? Different firmware, different label, different price. Western digital do the same thing, and I'm sure the others do to. Hitachi are probably allowing Apple server firmware, but won't let them have the Ultrastar label without paying the extra.
3. The only other "server grade" would be the 10K and 15K RPM much lower capacity, noisier, hotter drives designed for a continuous 24/7 workload. Using one of those would be silly.
4. Glad to see everyone's minutely interested in Apple still. I bet Dell & HP use the same drives in some server models, but nobody cares.
Disks are disks, in the end
I have a farm of about 200TB of disk in a variety of arrays, from EMC, DotHill, Pillar and others. The drives range from 7200rpm SATA, in the one of the EMCs, the Pillar and some OEM Adaptec storage, through 10K SCSI in the elderly DotHills and on through 10K FC in the remaing EMCs. We don't do 15K because an 8-way assemblage of 7200rpm is enough performance on the read side and our writes are bursty enough that the battery-backed RAM on the arrays isolates us from write performance.
The reliability appears totally unconnected to the brand and specification of the storage on the sample sizes I have. The Pillar has 100-odd 500GB SATA spindles, and (reaches for large pieces of wood) hasn't dropped a spindle in 18 months, for a truly awesome MTBF. It supports a large Clearcase environment and a whole bunch of Oracle databases, so it takes a pounding. The EMCs, on the other hand, have suitably fairy-dust sprinkled FC drives, and at one point we were on first-name terms with their maintenance people. That, and the hideous botch on Clarrions where the failure of one of the leftmost disks causes the write cache to stop working, made for some entertainment for a few months. We'd had a bad batch: of course you tend to get all the disks in an array you buy from a single batch.
The fun with the EMC SATA is that rebuilding an 8+1 RAID5 group to a hotspare with a centralised RAID controller is Not Quick: the Pillar improves on this (we've tested it) by having a RAID controller on each shelf to it can saturate the hot spare, but nonetheless rebuilding 500GB of parity is not rapid.
But for the Time Capsule, the whole thing seems moot. Unless you have 20 small laptops, the duty cycle for a Time Capsule will be a spin up, some writes and a spin down a few times an hour. In fact, a disk designed for 24x7xpounding may not be the best bet: a drive optimised for frequent spin down, like a laptop drive, might be a better fit. Six 120GB SATA 2.5" laptop drives in a RAID6 assemblage would be the thing, wouldn't it?
The way Time machine works
Try to not forget that if you leave it running on the dafaults, Time Machine will backup every hour of last 24 hours, then every day of last month, then a weekly one, until backup disk is full.
Since a full backup is done on first time then only changes are recorded,
if they are no changes since last backup, there's not much to copy either.
Then, there's few chances the backup drive will ever work 24/7, it will mostly idle.
Time Machine is for home users that will more likely use it because they accidentally earased some files than for pros who have real sensitive datas to be safely stored in 3 different physical places with UPS in each etc...
And since that one is working on wifi, you can even have it in an OTHER room of your flat than your computer after all !
RE: Wait a minute...
Well nice to see one from you though :)
some things never change
Or even the bloke down the road with an open WiFi connection and a time capsule - physical backups in multiple locations.
In fact, thinking about it, I might just set one up myself and see what gets left on it. One way to build up your p0rn collection.
Desktop Grade, good thing?
As opposed to Laptop Grade, remember all those Seagate 2.5" disks failing in MacBooks, not so long ago?
Whenever anyone says "server grade," I immediately think SCSI, to be honest. Despite the fact that my home server runs 2x500GB SATA-II drives, and my desktop runs 2x140GB 15krpm SCSI-320 drives... But that was down to capacity for the server and speeeeeed for the desktop.
Apple can't just go waving around phrases like "server grade hard drive" and not expect people to pull them up on it. "Server grade" doesn't refer to home servers, or hobby boxes - that's a phrase you use when talking about big, fat, SCSI/SAS/FC-powered bad boys. And as the Time Capsule is a home box, they should just say it's got a reliable hard drive.
Oh can we leave the Deathstar thing alone?
So one or 2 models of IBM's Deskstars had issues, and were able to be mapped to an amusing name. That doesn't mean the current series of Hitachi drives will do the same.
I've been running Hitachi drives in pretty much everything for years, and in that time, 1 has report errors (not actually died, it just reports SMART errors). That one drive got replaced because it was in my server (it's mirror partner went into my desktop where it now holds Vista for me).
I do see a high rate of failed Hitachi laptop drives, but I have a feeling that could be due to them being quite popular (almost all PATA laptop drives I've seen have been Hitachi).
In the end it's a moot point. Almost no one would pay for the thing if it had true server disks, and those who did would complain about the noise.
It just works!
... and for most people who use Apple computers that's all that matters.
I dumped my home Wintel kit 3 years ago after working with DOS/Windows for 24 years, and replaced it all with a Mac Mini and an iBook both now running Mac OS X Leopard. I use Time Machine/Time Capsule because I don't have to think about backing up my systems - ever again. I am a Unix SysAdmin/Consultant at work and don't want to have to think the same way at home.
But according to the reviews I've read of this device, it's hasn't even got NAS capabilities, it's purely a backup drive. So the unit doesn't even act as a real server and won't be under a heavy workload.
What Apple do to "Server" drives...
I think that "server" refers to Apple's selection process for the drives. When these drives are delivered to Apple they put them on soak for a few days and reject any with bad blocks - that's ANY errors. They then flash them with their own special firmware that allows for better error reporting and correction.
These error free drives are the ones they ship with XServes (and until recently) XServe RAID units. This is why a £90 drive costs £300 in Apple servers, not because they are fleecing us (well maybe) but that they have gone through the testing process.
I have had two fully loaded XServe RAIDs and 6 Xserves running continuously for the last 4 years. Apart from software updates, they have had 0% downtime. They must be doing something right
Those rejected ones go into the consumer machines - there's nothing really wrong with them, they just have some bad blocks.
Careful, you're in danger of using facts in a fanbois based argument.... you should know that is not allowed!
@Dan - i'd be interested to see what reviews you've read that say it's not a NAS device, exactly what else would you call a hard drive stored not in your local machine, but attached directly to your network and available to access remotely by machines?
Cutting to the real issue...
It's hardly worth quibbling about whether a server drive is more reliable than a desktop drive, did any of you actually read and comprehend the google paper on desktop drive life expectancy and MTBF in an enterprise scenario? If you had, you would've seen that desktop drives are just as reliable as posh server drives as Ian has also testified to a couple of comments above - this makes them ideal for usage in a high density low cost storage array. If you pick the right drives and controller they can also be very fast.
What is worth quibbling over is the fact that the unit is billed and advertised as having an enterprise grade drive and surprise, surprise it doesn't. It would likely be just as reliable whether it did or it didn't, but they have lied to market the thing. In my opinion the issue/news here is not the drive type used (which a lot of you seem to be quibbling over), but the fact that you have been lied to again by a large monopolistic company.
But Dan, it's sold as a back-up device - so why the but?
You don't WANT a "server" drive for this application...
Go take a look at StorageReview.com, and read their benchmark tests. The fact of the matter is, most "server" tuned drives perform much worse under single user loads than their desktop mirror models (such as this Hitachi and the Ultrastar server-clone). The firmware in server-drives is tuned for large numbers of pending I/O requests, and they scale a lot better when 4-8 people being to hammer them than when one person is using them. So for a single-user, low pending I/O load backup drive, the desktop drive will almost certainly outperform it's "server"-tuned cousin, and at a lower cost.
Moreover, many server drives are tuned to give up on time-out pending read requests in a way that makes sense when used in RAID configurations, not when used as a single disk. That helps their performance under RAID, but can allow them to give-up reading data in a single disk mode that CAN still be accessed if given time.
Reading through all the comments in this thread, I'm just really suprised how many people seem not to know this stuff, but claim to work in IT and read El Reg...
N.B. - those Hitachi's ARE good drives, I have a couple arriving soon for my own desktop system, which is on 24/7 and is a media server...
@Cutting to the real issue...
it pays to read the blurb Martin. Nowhere on the Apple publicity does it say "enterprise grade" so don't go making things up to bolster your argument that Apple lied.
The hitachi blurb says the drive is for use in network storage servers. So once again Apple did not lie.
Quibble about things that are misrepresented certainly - but all of your second paragraph quibble is misrepresentation.
The 'deskstar' is as it's name implies, designed for desktop use. That means on for typically 8 hours a day.
'Server' drives are designed to run 24/7.
TimeCapsule is clearly designed to be left on 24/7. I have no idea if it powers down the drive when not in use, but I would be *extremely* wary of trusting important data to a (particularly IBM) desktop drive. Before you argue it's a backup and you have another copy on the live system, lets pretend that you just lost your live system or want an old backup recovering. Still not important?
@ David Wiernicki - find your granny spectacles
Time to put on your granny glasses Wiernicki, as usual my truthful insights on the quality standards and the ruthless greediness of Apple's product managers was posted early Monday at "Apple has ALWAYS Cheated on Quality Components! "
I always new Apple Kool Aid Drinkers were stupid and in terminal denial, I had no idea that they were this blind too.
As a footnote to all this discussion - ANYONE who actually believes that Apple "designs" anything, rather than in reality finds a cheap supplier / manufacturer in the pacific rim with an already engineered product that will slap Apple's name on it, also believes in the Tooth Fairy and Santa Clause.
RE:Hitachi Deskstar and RE:server drives and server grade drives
Are you guys idiots? First of all Western Digital being as bad or worse then Fujitsu???????????
I am running a desktop with 4, yes 4 Raptor HDDs in it in a RAID 0+1 and i accidently broke the SATA connector for one of the drives off trying to hook it up blind and called WD for shlts and giggles to see how much a replace controller board would cost instead I waited NO TIME AT ALL to get someone on the phone after I made it through the IVR and once I told the guy on the other end what *I* had done and what I was looking for he said they would WARRANTY IT OUT!! How many other drive manufactures would do that? 2 weeks later brand new Raptor in my hands and in my computer. Problem solved and all it cost me was roughly $20 in S+H.
Now for the other knucklehead:
"It's hard ot find these beasties larger than 130 GB ... they invariably use 1.8 , 2 inch or 2.5 inch platters."
Erm say what?? Hard?? again 4 Raptors at 150GB each. And I believe they are 3.5 inch drives could be wrong though but I believe this link says it all:
Those look to be roughly 3.5 inches to me.
Next time think (and do research) before you open your gab.
/hate people who are idiots and talk before they think
- It's true, the START MENU is coming BACK to Windows 8, hiss sources
- iSPY: Apple Stores switch on iBeacon phone sniff spy system
- Chinese gamer plays on while BMW burns to the ground
- Pic NASA Mars tank Curiosity rolls on old WET PATCH, sighs, sniffs for life signs
- How UK air traffic control system was caught asleep on the job