Alternatives to Drobo Pro?

[Petri Teittinen]

First, I disagree that RAID isn't "a backup." That's exactly what it is. Has been for years.

Ask any IT professional and they'll tell you a RAID is not a backup. Ask any company selling RAID/NAS systems and they'll tell you the same.

One more example: copy the data on a HDD, then store the HDD in a clamshell box and place it in a secure container. A year later take the HDD out and connect it to a PC/MAC. In 15 minutes you have a lump of inert metal because the tiny ball bearings inside the HDD have fused together, caused by their lubrication drying out.

A RAID is the absolute best backup solution for data, I really don't see how that point can be argued. This isn't 1995 where you couldn't really rely on hard drive space because you could buy various tape media that was a lot bigger. This is 2009, almost 2010, where hard drive space is cheaper all of the time and there is quite a lot of it.

I think we might understand the term "backup" slightly differently. RAID might be the best way to safeguard the data while you using it, but once you're done with the project and wish to store it for a longer period of time (i.e. backup), RAID is not the way to go. Pixar stores their backups on LTO tapes, and I've been eyeing an LTO drive for a while now.

The odds of 2 drives failing within a day or two or even a week are extremely small.

This has happened to me three times within the past five years. RAID-5 every time, all data lost every time. Thank goodness for affordable RAID-6 NAS boxes.

The worst hard drives I've ever used personally were Maxtor. They were completely unreliable.

Damn, that brings back memories. Almost exclusive bad ones. Yeah, I agree, they were utter crap. And now Maxtor is part of Seagate...

 

[Hans von Sonntag]

The worst hard drives I've ever used personally were Maxtor. They were completely unreliable.

The only data media which never let me down is LTO from any vendor.

Hans

 

[Scott Brown]

I think a good solution is to have three lines of storage:

1. Frontline - this is used for you editing and needs to be fast, we use Raid 6 but could as easily use raid 0.

2. Second Line - Speed is NOT a priority here and solutions like Drobo come into play. Second line includes full backups of ALL media being used in your Frontline edit + projects that are waiting to be archived or are still current. Here Raid 5 or preferably 6 is essential.

3. Archive/Backup - no doubt in my mind that here LTO rules supreme. Hard drives are just NOT reliably enough. HD option may be ok if you have original sources tape but if you are in the tapeless world, then you need something robust.

Off site backup is another thing to consider - helps you sleep better at night!

Scott

 

[Jeff Kilgroe]

Jonathan, to be honest, I wouldn't touch Hitachi drives with a 10-foot stick. Simply because Hitachi HDDs are relabeled IBM DeskStars, a.k.a. DeathStars.

Ummm.... Or was it the other way around? Hitachi and IBM drives were essentially the same units from the same factory, cooperative design and construction. IBM and Hitachi were linked at the hip and in many ways they still are. Current Hitachi products are still a product of ongoing design at IBM and Hitachi.

Hitachi are the best ones out there today, shocking I know. Western Digital have been really good with their Green and Black lines over the past year or so as well. Seagates, I have had more bricks from Seagate in the past 2 years than I can count. But Seagate currently makes the best 2.5" HDD's, IMO. Actually the Hitachis are great too. I haven't had a failed 2.5" HDD in over 3 years. <EDIT> One failed in a RED Drive, but I think it got dropped<EDIT> The mobile drives are excellent these days. I had a crate of 1TB Seagates about 18 months ago where one out of every 5 was either DOA or failed within 90 days. That sucked.

The sad thing is that over the years, all drive makers have released product lines that were junk and all of them have released product lines that have been excellent.

I also have to warn you about placing too much faith on RAID arrays, especially RAID-5. I have several years of experience with both self-assembled RAIDs as well as NAS boxes, and I've learned my lessons the hard way.

110% agree. And I see too many people on these forums thinking that they're secure by placing their data on a RAID-5 or RAID-6 volume. No storage is totally safe and secure. Data security comes through redundancy and proper management. I advise multiple copies on various types of media. RAID is great for near-line storage, live backups, active workspace.

The thing is, when you buy a pile of HDDs, whether inside a NAS or from a retailer, the HDDs are usually from the same production batch. This raises the odds that the pile might contain drives that will fail around the same time.

Yep. See it all the time. One of the last RAIDs I bought was a CalDigit HD-Element with 4x1.5TB Seagates inside. One failed within a month of buying the unit, I replaced it. Another failed shortly after and I replaced it, a third one failed while the RAID unit was rebuilding the array -- all data lost. I had a backup, but still lost about 3 days worth of work. All drives have been replaced in the unit with more 1.5TB Seagates, but with updated firmware that supposedly is less aggressive with head operations and all has been well now for over 6 months. If another one of those drives fail, I will swap all 4 out with 2TB Hitachis. The 1.5TB drives seem to be performing well out in the field now.. But with 1TB and 1.5TB Seagate released several runs into the market with faulty firmware that caused the drives to brick themselves under certain conditions. They even recalled several versions of their 1TB HDDs due to this. It was either that or face class action litigation.

For those looking for a low-cost tower RAID, take a look at the SansDigital unit with SAS interface. It has two mini-SAS connectors, no multipliers and gives 300Gbps per HDD. They throw in a SAS controller too, but I'm not sure which one. You can buy it bare with no drives for about $1200 with the controller. For $2400, you get the controller, cables, and 8 x 2TB HDDs. It's a nice and fast system on the cheap. But it doesn't have redundant power and all the same bells and whistles of some of the more expensive RAID systems.

Try to avoid building desktop or deskside RAID systems with eSATA. Most systems sacrifice bandwidth by using only one or two SATA cables for connectivity to the RAID and subsequent multipliers inside the enclosure. Then again, some of those options get really cheap.

SansDigital has some good mobile options too, but I haven't worked with them myself. For mobile RAIDs I like the CalDigit VR (for a 2 drive RAID), they're a bit more expensive than G-Tech, but a lot better.

 

[Jonathan L. Bowen]

Alright

Alright

I've given this a lot of thought over the last day, because you are all smart people, and I did some more research and thought about why I believe what I do, and why that differs from some people here. It comes down not to a difference in opinion but a difference perhaps in what we're doing with our data, what value it has to us, and what our businesses maybe do primarily.

I thought, "Well, I haven't looked at tape backups in a while, let me go see what is new in the world of tape backups, as I've never heard anyone talk about them since the late 90s." Alright, so I go and try to find some solutions, what are people using, etc. I find Iomega has a decent solution with 35 gig backup tapes essentially, they are $69 per one. Then I see this LTO and other stuff, bigger capacities like 350 gigs even I believe on one of these tapes. Not bad!

My problem with all of that is it's not only much slower than a RAID array, therefore useless for editing (and of course you'd say, well yeah, it's archival purposes, not editing), but it's also overly expensive for what my data is worth. I don't think we disagree that tape is a great method for backup of really expensive data over a period of 30 years or something along those lines. I think we disagree because the cost-benefit analysis makes a tape backup a terrible solution. Worse than that, really, it's totally impractical.

Let's say that a client commissions my company for the absolute lowest package we offer, which is a $1,500 virtual tour. For that, I don't direct it, I don't really do anything but hire a videographer, hire an editor, then the videographer shoots according to the specs I give them from the client, the editor does what I want them to do with the footage, I don't really oversee any of this besides giving a few pieces of advice. I spend maybe let's say $750 total on both people plus my insurance cost and sales person commission. So I take home $750 for being "quality control" and customer-relations management, basically making sure the client is happy. It's reasonable to expect that shooting with an HVX200, the person could go through maybe 20 to 25 gigs on a simple project like this. Perhaps only 15 gigs, who knows. But still I'm going to be left with saying alright, I want to back this project up twice, once perhaps on my DroboPro RAID 6, and then I want to throw all of the stuff onto some external drive as emergency backup. Now I guess a 1 TB external goes for what like $100? $110? I don't really remember, been six months since I was looking at externals. Mine are all 500 gig externals from 1.5 years ago or so. Well a tape backup solution, just for the tape, is going to cost me let's say $50 to $75. That doesn't make economic sense for a project of this type. My profit margin is too low to be spending 10% of my profits on backing up the data.

Now let's say I shot a feature film for $2 million and I really want to make sure my data is secure. I would probably take the footage and put it on a DroboPro in Los Angeles, I would probably buy a second DroboPro to put in Portland, OR, where I have space as well and not only that could get people to walk by it and make sure those green lights are on every few days. I could then spend $10,000 on tape backups without it being a big deal, figuring eh, you know, $2 million compared to $10,000, no biggie!

But with the type of work I primarily do, which is corporate work between $1,500 and $6,000 per project on average, let's say $8,000 to $15,000 per training video or whatever, it's simply not high enough budget work to make those additional security measures necessary.

All I can say to the guy who lost data three times in five years on RAID 5 systems is -- ouch. You have to understand that's not normal. I mean, if you look at the statistical failure rates of hard drives, a RAID 5 is a very secure solution for storing data. It's not perfect at all, it has vulnerabilities, and two failures can happen. It's incredibly rare across the population, so basically for you, it's just been terrible luck compared to most people. I've had in my life I believe 3 hard drives fail, 2 were Maxtor, 1 was Seagate. That's not too bad, I guess, I've probably owned 50 hard drives, just guessing. Obviously RAID 6 is a more secure solution.

What I'm trying to say is sometimes how you treat your data or the measures of security you go through depends highly on what that data is worth and what makes sense as a businessman. You are going to treat a small corporate project differently than a feature film. Even for my first feature that I want to shoot next year, the idea has occurred to me that it would be nice to have a second DroboPro in another state where all of this data can be backed up as well. Honestly I have to say if that doesn't give you peace of mind, though, you've really had some bad luck. If I have a feature backed up on two DroboPro units in two states, then a third backup on external hard drives in storage here (on a different floor, thus different physical location but same building), honestly I feel very secure and happy with that. I wouldn't want to spend $10,000 more to put it on tape, that just makes no sense to me. My level of paranoia for a $150,000 feature is not that high, I guess. I think some people want to back up data to a medium like tape where they feel it's not going to fail on them, and they don't ever have to check it. I'm not sure I really think that's as secure as you think it is. First, I would say if you have it on tape, and not double-backed-up to tape, your solution is far less reliable than RAID solutions. FAR less. A RAID unit like DroboPro or CalDigit HDPro or HDOne all have sensor lights on those hard drives and RAID software programs -- they will tell you immediately when a drive failure occurs. You can even get them to e-mail you, for instance, as was mentioned here. We have the same setup for our co-hosted server, where it e-mails us of anything going wrong with the server itself.

The problem with tapes is you need a lot of them, or at least a decent number, and you need a separate hardware device (at least I've seen these) that verify the tape's functioning order still. That still requires someone to do that task every six months or year, or whatever you want to say. Then there's the whole issue of if you think you can leave tapes alone for 30 years and it'll all be fine. Not so sure about that. Will the same tape drive technology even exist then? I guess you could find an old tape deck that will work. But hard drives haven't changed in 15 years -- they are bigger, yes, they are faster, yes, but they're still 3.5" HDDs, they are essentially always the same type of thing just bigger and faster. It's actually quite reliable. They're probably not going to suddenly change to a 5" hard drive, for instance. I personally think it's a more time-efficient solution to have a RAID unit where you have a simple light pattern showing you the drives are active. If you're noticing that on a daily basis, you're going to know very fast if a drive fails. You don't even need to open software for that. You could argue, "Well what if the lights don't work right." Look, enough paranoia over anything is going to make you worried. As I said if the data is that valuable, totally, you should have it on everything you can get it on -- RAID, externals for emergency (double-backed-up), tapes, etc.

I am merely saying that for my business that makes no sense. The cost would be prohibitive. And I can tell you from talking to several big computer nerds in the last day, none of them agree that a RAID isn't a very secure backup solution. Also whoever said ask the manufacturers, uhh is this a game of chicken or are you serious? Just curious because you really, umm, don't get it! Go to Drobo's Website, look up their advertising materials, their news articles, their video contest -- ALL emphasize that your data is secure against data loss with DroboPro, that's their big selling point. Numerous video companies have sworn by them, for instance. Go to CalDigit.com and you'll see "fast, reliable, secure" data solutions. They sell themselves on offering a very secure solution for data storage. That's what RAID is all about.

If you think RAID is about just fast editing, you're wrong. I shouldn't have to tell a media professional this. The DroboPro isn't even an editing solution, it's a storage solution. It's only Firewire 800, which is pretty slow. The HDOne is very fast and it makes sense to get that just as a fast-editing solution, which is in fact why I got mine, not for backup of important files, which I was just doing with multiple externals at the time. Having a RAID backup is a definite step up from that. I'm not denying that on top of that having a tape backup solution as well would be *even better*, but I'm just saying that only makes sense for some projects.

Also Jeff is right, I had looked more into the whole "Deathstar" drives, and that was basically 2002 / 2003, it sounds like, that is quite a long time ago in the data world. I have since read a lot of really positive reviews for Hitachi being excellent quality drives now. I think every company makes a total dud line at some point. Seagate lately has gotten a bad reputation, for instance, but for a while it was considered the most reliable drive solution. My reason for being a huge Western Digital fan is simply because of ten years of personal experience with buying a ton of their drives and having 0 failures. It's not any more scientific than that. It could be purely great luck, lol. I know for instance that the 2.5" Seagate 80 gig drive installed in the Macbook Pro models in 2006 was a total dud -- such a dud that a class-action lawsuit was filed against Apple for not recalling their laptops after admitting "there were problems" with those Seagate drives.

 

[Jeff Kilgroe]

I thought, "Well, I haven't looked at tape backups in a while, let me go see what is new in the world of tape backups, as I've never heard anyone talk about them since the late 90s." Alright, so I go and try to find some solutions, what are people using, etc. I find Iomega has a decent solution with 35 gig backup tapes essentially, they are $69 per one. Then I see this LTO and other stuff, bigger capacities like 350 gigs even I believe on one of these tapes. Not bad!

For tape backups, the best game in town right now is LTO-4. Either HP or Quantum. The HP is preferable by many and has much better compatibility with various SAS and SCSI controllers, which makes it better for OSX and Linux. LTO-4 tapes are $40 each and hold 800GB uncompressed. You will need to buy additional software for proper operations of these tape systems. They don't simply connect and work as a hard drive. Transfer rates are good. Desktop LTO-4 units can burst up to 120MB/s, but sustain around 80MB/s in most backup operations. They tend to read about the same speed. Some of the larger LTO-4 mechanisms can double those transfer rates, so in that respect, all the LTO-4 options are better than FW800 in most cases. Of course, tape is a linear system, so those transfer rates can only be sustained while the tape is reading in linear fashion off the tape. It's possible to extract single files from tape, but once again, you can't use them as a hard drive.

The biggest downfall to LTO and most tape backup systems is the initial buy-in cost. Figure about $3.5K to set up a desktop LTO-4 system with a handful of tapes to get started. From there on, tape is typically a bit cheaper per GB than hard drives and the initial buy-in cost is usually recovered after you archive somewhere around 25-30TB or so.

In today's world of data, ideal solutions typically involve RAID systems for both high-speed workspace as well as near-line storage and backups. LTO tape for archival. LTO is the leading industry standard the world over for long-term data archival media.

There are lots of us here on reduser who are archiving to LTO.

It's reasonable to expect that shooting with an HVX200, the person could go through maybe 20 to 25 gigs on a simple project like this. Perhaps only 15 gigs, who knows. But still I'm going to be left with saying alright, I want to back this project up twice, once perhaps on my DroboPro RAID 6, and then I want to throw all of the stuff onto some external drive as emergency backup. Now I guess a 1 TB external goes for what like $100? $110? I don't really remember, been six months since I was looking at externals. Mine are all 500 gig externals from 1.5 years ago or so. Well a tape backup solution, just for the tape, is going to cost me let's say $50 to $75. That doesn't make economic sense for a project of this type. My profit margin is too low to be spending 10% of my profits on backing up the data.

You're right that LTO or similar tape solutions don't make sense in that situation. 1TB drives are about $80 these days, 1.5TB are about $110, and 2TB are about $165.

But for small projects -- maybe you should consider optical media? Blu-Ray writers are less than $250 for a good external unit, under $160 for an internal. Blank 25GB BD media is $4/disc. There's DL 50GB media, but they're too expensive by comparison - $14 or so in bulk as of a couple months ago.

Also Jeff is right, I had looked more into the whole "Deathstar" drives, and that was basically 2002 / 2003, it sounds like, that is quite a long time ago in the data world. I have since read a lot of really positive reviews for Hitachi being excellent quality drives now. I think every company makes a total dud line at some point. Seagate lately has gotten a bad reputation, for instance, but for a while it was considered the most reliable drive solution. My reason for being a huge Western Digital fan is simply because of ten years of personal experience with buying a ton of their drives and having 0 failures. It's not any more scientific than that. It could be purely great luck, lol. I know for instance that the 2.5" Seagate 80 gig drive installed in the Macbook Pro models in 2006 was a total dud -- such a dud that a class-action lawsuit was filed against Apple for not recalling their laptops after admitting "there were problems" with those Seagate drives.

Western Digital had a terrible run of quality control across the board back in the early '90s. They faced bankruptcy at one point. They also had a bad run on a few different drive models here and there since. Back then, the hard drive manufacturer landscape was quite different. There was IBM/Hitachi, Micropolis, Panasonic and Kyocera were in the HDD business, JTS in the late '90s... But bad drive models happen with any of the manufacturers. Lately Seagate has had most of the bad luck it seems. They are still a great company and overall produce great products.

Anyway, going back to drobo... Their solution is more about convenience than anything else. Think of drobo as "RAID for Dummies". It's a very solid product, but the convenience factor does create a compromise in both performance and data efficiency. They are excellent products and seem to be very reliable, if you can live within their technical limits.

I have used a drobo pro a bit, played with it, but not extensively. I do own a regular drobo. Nice little unit, slow, not the most efficient use of space, nor the most robust RAID implementation, but reliable thus far and it couldn't be simpler to use.

 

[Tom Lowe]

I looked at tape, and now I'm starting to think that Bluray will just be a whole lot more convenient for secondary backup of Drobo/RAID arrays. I can mail them home from the road, etc. Cost wise, they're great.

 

[Jeff Kilgroe]

The only issues with Blu-Ray is the cost per GB and you will pile up a lot of discs really quick. It's quite high compared to HDD or tape. But the cost to buy-in is like nothing. BD media is about $0.16 per GB on 25GB discs. LTO-4 is about $0.05 per GB, but you have the higher buy-in costs.

If you compare 10TB of backup data, that equates to:

13 x LTO-4 @ $40 + $3500 for kit & software == $4020

400 x BD-25 @ $4 + $250 for external kit == 1850.

So, at 10TB, Blu-Ray is about half the price of LTO-4, but a lot of physical media to manage. Blu-Ray is also SLOW!!! Think USB2 speeds in most situations.

Tom, if I understand what you're going to be doing, mostly timelapses and whatnot, Blu-Ray seems to make a lot of sense. Either that or a small RAID-1 ecnlosure where you can make duplicate HDD's of your footage. No matter what format you choose, I would recommend making dual backup copies.

 

[Tom Lowe]

Thanks for the advice, Jeff.

BTW, Drobo finally has eSATA! http://www.drobo.com/products/drobo-s.php It's only a 5-bay Drobo S, but hopefully they will add eSATA to their 8-drive as well.

 

[Jeff Kilgroe]

Hey, that Drobo S just popped up, uh like today. Don't see a price for it yet. 5 drives, dual drive redundancy. That could be a real winner if it can saturate an eSATA bus.

I'd love to abuse one and see how it handles if I can get my hands on one.

 

[brandon thomas]

i'm sure the eSata will speed up the drobo, but it won't reach speeds anywhere near a typical eSata device. I think the speed hit comes from the mix/matched drives and double redundant structure they've built upon.

I know from experience the FW800 Drobo is painfully slow, and hopefully this is something they've greatly improved upon in these new versions.. but i doubt you'd be able to comfortably use any Drobo as a primary storage device..

 

[Jeff Kilgroe]

A single SATA 3Gbps connection is hardly fast these days. I'm hoping that with 5 drives, they could get close to 300MB/s or so when in single-drive redundancy mode. Not a big trick for any decent hardware RAID-5 these days with 4+ drives.

The current 4-drive Drobo, I own one, is painfully slow. But with decently fast matched drives, it can still saturate a FW800 bus with plenty of speed to spare, just the interface can't cope. I'm tempted to pick up a Drobo S and put 5x2TB 7200rpm HDDs in there to see what happens. I'd like to think that is should get close to the SATA limit. But I'm typically a "glass is half empty" sort of guy, so will concede that you are probably right and that the performance will suck, even more than one would expect it to.

 

[Martin Weiss]

Drobo just introduced two new models, the Drobo elite and the Drobo S.

The Drobo Elite looks pretty much like the Pro, except that you can have 2 disk redundancy.

The Drobo S finally has an eSata port, so it will be interesting getting some real world speed tests from that baby. 5 disks.

Edit: next time I'll read all posts before posting

 

[Jeff Kilgroe]

Saw the updated pricing now that they have it online. Hopefully the Drobo S street price will be much lower than MSRP. With the prices on the Drobo Elite, why bother. OMFG... Stupid with a capital "S". Especially with the specials offered for us around here by MAXXDIGITAL and others.

 

[Chris Romero]

Why does Drobo continue to fill their units with "desktop" drives. This just shows they don't care about quality and just want to keep selling their junk.

Desktop drives or consumer class drives are just not meant for RAID environments. I deal with people all the time that have issue after issue because they decided to use "desktop" drives in a RAID to save a little money. Drobo is not the only company that is doing this, Caldigit, Maxx and Gtech are also guilty of this crime. Just spend the extra money now on "enterprise level" drives for your RAID and you'll have much less of a headache later.

Now, about Drobo allowing to mix drives in a RAID? Why? Why put a brand new 2TB drive with a 300, 500 drive? I understand you want to increase your storage, but rule #1 in RAID, never ever mix drives, they all need to be the same. The inside architecture of a 2TB drive and 500GB drive are so different, not to mention the firmware differences. It's just mind blowing what these companies are offering now...junk.

 

[Jeff Kilgroe]

You do realize that in most situations, the difference between "desktop" drives and those "enterprise" drives is just the label, right? Current Hitachi and Seagate "enterprise" models are identical as far as hardware goes, some models will have different firmware, but some use the same firmware the "desktop" models use. Typically what you pay extra for is the extra warranty.

CalDigit and MAXXDIGITAL both use "Enterprise" class drives when available. The only CalDigit product I've ever received that did not ship with "Enterprise" class drives was an HD-Element unit about a year ago that shipped with 1.5TB HDDs. Seagate never offered an "Enterprise" version of those drives.

Always best to research each drive model and see what's what. HDD manufacturer warranties are rarely worth much and the hassle of the process usually outweighs just buying a new HDD. Especially when you purchase large volumes of drives.

I'm not saying that one should not buy the "enterprise" class units, but sometimes there is no advantage or difference. Just research each drive model before jumping to conclusions. That said, I've had "enterprise" class drives fail on me just as often as "desktop" drives.

 

[Chris Romero]

Jeff,

I've work with Maxx, Caldigit and all the others for many years. The only reason they switched over to enterprise level drives was because of the continued issues they kept having with desktop drives. Maxx used to try and feed me the same story about there being no differences between desktop and enterprise drives. If so, then why the sudden change to all enterprise level drives now? I'll tell you why, because enterprise drives are so much reliable and build for RAID environments (vibration) and since the change, they experience far less time out errors and drop drive issues. So you can believe what you want, but the facts are there. Enterprise level drives have far less issues than desktop drives.

I also deal with drive and RAID controller manufacturers, like WD, Hitachi, Areca, ATTO...they all agree about choosing enterprise vs desktop.

Now, there's always going to be that person that is looking for a quick bang for the buck...well, you can't go wrong with "desktop" drives...but these are also the same people that can't understand why their RAID keeps failing.

By the way, Hitachi does offer a "desktop" drive called the Deskstar series.

 

[Chris Romero]

By the way, I found these as good alternatives to the Drobo S.

G-Technology - GSpeed ES

Areca Technologies - ARC-5020

Also, I just talked to CineRAID last week. They will have a 4 bay eSATA unit and a 12 bay eSATA with a built in RAID. According to them, the 12 bay will only use 1 eSATA cable to support all 12 drives! Although, they said it will top at 200+MB/s (which is enough for some RED users), this is mostly meant for storage convenience and easy laptop connectivity. CineRAID still recommends using their editpro 4 bay with PCIe hardware RAID controller for top performance.

 

[Jeff Kilgroe]

I've work with Maxx, Caldigit and all the others for many years. The only reason they switched over to enterprise level drives was because of the continued issues they kept having with desktop drives. Maxx used to try and feed me the same story about there being no differences between desktop and enterprise drives.

I think you missed my point in there somewhere, about researching each drive make/model. And regardless of whether a drive is marketed as "enterprise" or "desktop" doesn't necessarily mean there is actually a physical difference. In many cases there truly is not, in some cases there is.

I purchase hundreds of drives in a year, close to a thousand last year, for various business operations I'm involved with. I've taken many of these HDDs apart, installed "enterprise" firmware on "desktop" drives, etc..

What I'm getting at is just because an "enterprise" drive is labeled as such, doesn't automatically make it better or more reliable. And in many cases it's no different than the non-"enterprise" version. Not all "enterprise" level drives are better tuned to operate in a RAID, assuming they are tuned differently at all.

And like I said above, about "enterprise" level drives failing, it happens. I bought a partial palette, over 100 UltraStar 7K1000's a few months back, and had several drives that were either DOA or failed within 90 days.

I've only bought a few CalDigit products, spread out over the past 3~4 years. They have always used "enterprise" labeled drives, except the one unit I mentioned. But Seagate still has no "enterprise" version of those drives. MAXX does not offer 1.5TB drives because of this reason. I don't know when they supposedly switched over to only using "enterprise" units, but I had no experience with their products until late last year anyway.

And yes, I know what Hitachi DeskStars are, but thanks for that.

 

[Tom Lowe]

Jeff, with the Drobo S (5 bay) now having eSATA and only costing 800 bucks, could I conceivably build a PC with four 2TB drives in RAID 5 plus one fast SSD for OS and programs (this PC containing all my hero shots and an ongoing rough cut of my film), then back up this PC RAID 5 array data to one or two of these Drobo S units for safety? Does that seem reasonable/plausible?