Stories
Slash Boxes
Comments

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

Sun Unveils RAID-Less Storage Appliance

Posted by ScuttleMonkey on Mon Nov 10, 2008 05:26 PM
from the testing-the-cutting-edge-of-data-storage-sounds-less-than-fun dept.
pisadinho writes "eWEEK's Chris Preimesberger explains how Sun Microsystems has completely discarded RAID volume management in its new Amber Road storage boxes, released today. Because it uses the Zettabyte File System, the Amber Road has eliminated the use of RAID arrays, RAID controllers and volume management software — meaning that it's very fast and easy to use."
+ -
story

Related Stories

This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More
Loading... please wait.
  • Pricing:
    Sun Storage 7110: $10,995 for 2TB;
    Sun Storage 7210 starts at $34,995 for 11.5TB;
    Sun Storage 7410: Single node version starts at $57,490 for 12TB;
    cluster version (with two server nodes) starts at $89,490 for 12TB.

    Correct me if I'm wrong, but doesn't charging enterprise prices for simplified hardware that relies on commodity software solutions, kind of defeat the point?

    Unless I'm misunderstanding this hardware, the entire idea is to move data safety away from hardware redundancy toward software-driven duplication. In that way, the data is safe from failure in the same way that GoogleFS protects against individual machine failures. The only difference is that Google probably doesn't pay $11,000 for 2TB of storage. :-/

    One of these days, I really will understand why Sun regularly shoots themselves in the foot. Until then, I suppose I must trust them to somehow find a customer who's willing to pay exorbitant prices for an otherwise good idea. (i.e. I'd really love to see Sun bring Google-style reliability from unreliability to the market.)

    BTW, here's the link to Sun's marketing on this:
    http://www.sun.com/storage/disk_systems/unified_storage/index.jsp [sun.com]

    It's actually pretty cool tech. Sun could own the market if they just understood how the market views pricing and features.

    • by poot_rootbeer (188613) on Monday November 10 2008, @05:28PM (#25712471)

      I suppose I must trust them to somehow find a customer who's willing to pay exorbitant prices for an otherwise good idea.

      Have you worked with any of Sun's customers recently? I believe P.T. Barnum was involved in the development of their business strategy.

      • by Anonymous Coward on Monday November 10 2008, @11:52PM (#25716327)

        This is meant to be 100x faster than the storage you're talking about:
        First [sun.com]: This uses Hybrid Storage Pool:
        The Hybrid Storage Pool combines DRAM, SSDs, and HDDs in the same system, dramatically reducing bottlenecks and providing breakthrough speed.
        Second [sun.com]: The system's hybrid architecture gives you the speed and performance you need to shatter the I/O bottlenecks with no administrator intervention. In fact, Hybrid Storage Pools with SSDs can improve I/O performance by 100x compared to mechanical disk drives.

    • Re: (Score:3, Interesting)

      > Correct me if I'm wrong, but doesn't charging enterprise prices for simplified hardware
      > that relies on commodity software solutions, kind of defeat the point?

      Yea, that is amazing. Ya could put in a pair of 1U servers with RAID1 on each for a fraction of that pricetag. Use any of a number of ways to make the two units cluster, including using OpenSolaris and you get everything they are selling except the pretty front end for about half the sticker, Go SCSI/SAS on all of the drives in 2U machines

      • by chrisj_0 (825246) on Monday November 10 2008, @05:49PM (#25712733)
        I don't think you could get the IOPs (or anywhere near) out of a pair of off the shelf 1u servers that they're advertising. I just checked dells website, their new AX4-5i (iscsi SAN) starts at over $14,000 and that only includes the 4x 750GB vault drives. Add 4x 1TB SATA drives (at $1,100 each) in a RAID 10 and you still wouldn't get the IOPs that Sun is talking about. This product looks to try and take a market share from the FC SAN vendors, not companies that want their in house geek to build a "cluster storage solution".
          • by Score Whore (32328) on Monday November 10 2008, @11:29PM (#25716181)

            Is that 2U 1M IOPS unit racked right next to your 1U, 64K core, 1024 TeraHertz system?

            FYI, a loaded HDS 9990V (i.e. hundreds of spindles and multiple gigabytes of cache) manages to provide 200,000 IOPS (SPC-1). Even Texas Memory Systems RamSAN 400 (i.e. SDRAM) can only make 400,000 IOPS. Hell, it was only a couple of weeks ago that TMS was announcing that they sold a RamSAN-5000, which is the only storage device I've ever seen specced to 1,000,000 IOPS. And it's 10 different RAM cached, flash backed units.

              • by Giant Electronic Bra (1229876) on Monday November 10 2008, @09:50PM (#25715343)

                to me. Coming from high performance transaction processing land where an operation means 'the data is ON the platter' you can't do that more often than the platter rotates to the point where the head is over the sector where the write operation starts. Basic math, 15k RPM spindle = roughly 300 times/sec. Multiply by however many spindles you got, that's what you're max throughput is.

                This is one reason why IN THEORY at least an SSD would be so great, that latency is much less. So basically I'm thinking they just aren't talking about what you're talking about, and maybe that makes sense, if you're running a trading operation say, you just DO NOT CARE what is buffered someplace, if it isn't physically on the drive, it doesn't exist.

    • by darkjedi521 (744526) on Monday November 10 2008, @05:33PM (#25712543)
      Some of that is the custom gear that goes into making those beasts. Yes, it might eliminate the hardware raid card, but in the case of the 7210, the hardware to drive 48 SATA drives and not saturate the bus still isn't cheap. Plus hotswap everything, and the price quickly rises to something close to what Sun is charging. I use 4 x4500s at work for a single cluster, and they are a hell of a lot cheaper for that capacity than the traditional rack of fiber arrays/raid controllers/etc. The 4 of them cost me what another vendor wanted for half the raw storage (and far less usable storage).
      • the hardware to drive 48 SATA drives and not saturate the bus still isn't cheap.

        If you're driving 48 SATA drives on one bus, you're:

        A) Not looking at the minimum 11.5TB layout
        B) Not paying $35,000
        C) Not a small-business customer

        Which brings me back to: Sun is promising to target the small business and yet totally missed the mark. This is Enterprise hardware.

        • by kandresen (712861) on Monday November 10 2008, @08:22PM (#25714565)

          Just a comment about the 48 disk setup; it is not always about getting the most space, but often about getting fastest response time. In this case the important factor is the amount of spindles. 11.5TB divided on 48 disks would be ~240GB a disk. Many companies would want 48 70GB disks as they are not in need of more space, only faster response times.

        • Re: (Score:3, Interesting)

          Small Businesses are businesses that make under $25M/year by definition. I can imagine small businesses being in the market for inexpensive, high throughput, SANs.

    • by rainer_d (115765) on Monday November 10 2008, @05:49PM (#25712747) Homepage

      I don't think it's that expensive.
      I use Promise's VtrakJ610s at work (16x1TB SATA), and it cost about half that - but I still need a server for it (DL385 in our case). And I need to fit the disks myself (16x4 countersunk screws...) into the ultra-cheap harddrive containers.
      A MSA70 full of SAS-disks (25) costs 10k, IIRC - but you need a server, HBAs etc.
      I'm soooooo sick of the "I could build one for XXX% less using YYY"-comments.
      Please, all the winers: go and start your own company selling and supporting storage-systems.
      Good night and good luck....

    • by sexconker (1179573) on Monday November 10 2008, @06:04PM (#25712927)

      It's SUN.
      They're goal is to stay relevant, their strategy is to make headlines.

      It's like a cross between a child acting up for attention and an emo cutting themselves.

    • Re: (Score:3, Insightful)

      the entire idea is to move data safety away from hardware redundancy toward software-driven duplication

      You are exactly right. When you pay the exorbitant price, you pay for great hardware, the development of great software (which you could have gotten for free), the convenience of a prepackaged solution, and for the hardware and software support.

      Should anything happen to these machines, you can always get your data back. If you can't afford another set of machines like these, simply plug the drives into anything that runs Solaris (or generally ZFS), and you have your data.

      Just because it's open doesn't m

    • Re: (Score:3, Insightful)

      • Sun X4240, cheap from third party reseller not necessarily in the best of configurations: ~£2700.
      • 2 cheap and cheerful SSD's for ARC second level cache: £300. £1200 if you want ones with decent write performance.
      • 14*146GB 2.5" 10kRPM SAS disks: £2100.

      Even if you put it together and test it using slave labour, you're not getting much change from $11k.

      Sure, you could just plonk three 1.5T Seagates in there, shove a RAIDZ over it and call it a day, and that would

    • But SUN is FAR from being the inventor of charging people $50k for something they could just as well get for free...

      Name ANY big IT vendor, they all do it. My father can tell some amazing stories on that subject. Not a new phenomenon either.

      Now, if you are the GOVERNMENT, they'll give you the special bonus public sector price, $150k!!!

      • by Famanoran (568910) on Monday November 10 2008, @05:56PM (#25712815)

        With the same level of assurance that the solution will operate, first time - every time?

        With the same level of confidence that Some Vendor will bend over backwards to fix it if it doesn't work?

        Will your solution be as well tested and engineered?

        It's not like you can just grab 3 1TB SATA drives, throw them into RAID-5 and say that you've got 2TB of production ready storage. Well, you can, but you'd be an idiot.

        Your "home brew" solution will not meet any of the objectives Sun are achieving with this product. Your spindle count will suck, so concurrent access will be slow. You will probably be limited to one of iSCSI, CIFS, NFS or WebDAV, I doubt your solution would have all - and if it does, the integration will suck.

        Will your solution have the diagnostic tools that Sun can provide? Oh wait, you don't have the millions of dollars to invest in engineering quality diagnostics, right from disk analysis (Sector scanning, remapping, etc) through to performance related faults? Well, then your solution will suck. What about snap-cloning?

        In short, yes - storage is cheap. You can grab large drives very cheaply and put together something that works. That does not mean it will be good. Production quality storage is expensive, and for good bloody reason.

        As for doing this using SSD storage, that's just ridiculous. 2048GB of storage would be at least 16 128GB SSD disks - this is not counting any disks for redundancy (i.e, raid-5/6 parity), or hot-spares. Assuming 2 drives for RAID-6 parity and 2 hot swaps, you'd need 20 SSD disks - with 10 grand, you're expecting to pay $500 per disk - and no other hardware, i.e, motherboard, case, cooling (more important than you think), etc.

        So, until you have a clue about designing production quality storage systems, please refrain from making statements you have no clue about, you're only serving to confuse those people who are actually interested in what this product has to offer them. Keep to building crappy 3 or 4 disk RAID-5 systems using extremely large drives for storing your music, movies and pr0n on, but don't ever ever ever ever think about using those in any situation where your financial livelihood depends on that data.

        • by Archangel Michael (180766) on Monday November 10 2008, @06:24PM (#25713167) Journal

          Okay perhaps not with SSD.

          I've built 1.5 TB systems over a year ago, RAID for $1200. FULL including swappable drives, gig ethernet and plenty 2 GB of ram to cache the system. They are FAST and reliable. They worked FIRST TIME, and have worked exactly perfectly for 1.5 years. Drives fail, we have hot swaps available. While not quite 2 TB, they are also 1.5 years old now, and I'll be replacing them in another 1.5 years with bigger drive systems.

          And these, will be spares and lower priority sytems when I update them with newer stuff in a year or two.

          Expensive Technology for the sake of all that other stuff you listed is just silly. It is exactly why SUN doesn't get it, and why some pointy hair boss is buys the bs.

          Production quality storage means that it works for the time needed. I've actually had WORSE reliability from Name brand "Server" quality stuff. We've got HP Proliant Servers in production, and at least THREE from three different lots have all failed due to MOBO Failures. While they do send out a tech to replace the MOBO, it is really really annoying to have to tell people that the server is down because the MOBO failed. And all the great diagnostic tools HP has on those servers didn't predict nor would they fix the errors.

          You can build it 1/2 as much then you can easily have two on hand, in case one dies.

          • by Architect_sasyr (938685) on Monday November 10 2008, @06:50PM (#25713503)

            We've got HP Proliant Servers in production

            Some things you should keep to yourself no matter how bad it operates ;)

            This is a real case of quality, support and "bling" factor. To use a (bad) car analogy: There is no need to buy a Mercedes when you can own a Nissan for half the price and it has exactly the same features (it may even be more powerful in some cases). However anyone can drive a Nissan (or can afford to), so there is a certain bling factor to driving the Mercedes. Just like there is a hell of a "bling" factor to owning Sun equipment as opposed to the "hack job" we can all put together. Personally I would prefer to spend twice as much and know that it's no longer my problem, even if it crashes, but that's just the opinion of one Network Admin.

            Completely off hand: I've never had a mobo fail in any server, IBM or Dell based.

        • With the same level of assurance that the solution will operate, first time - every time?

          Sure.

          With the same level of confidence that Some Vendor will bend over backwards to fix it if it doesn't work?

          Heck, I'll even throw in the same vendor!

          Will your solution be as well tested and engineered?

          Even better. It will have had the same testing and engineering, PLUS a pre-existing history of operating in the marketplace.

          I give you, the Sun Fire X4500 Server [sun.com]:

          12TB (48x250GB) - $23,995.00
          24TB (48 x 500GB) - $34,995.00
          48TB (48 x 1TB) - $61,995.00

          Let us compare with Sun's new line, shall we?

          11.5 TB (46 x 250GB) - $34,995.00
          22.5 TB (45 x 500GB) - $71,995.00
          44.0 TB (44 x 1TB) - $117,995.00

          So... twice the price for the same storage? To steal a line from a very famous "programmer":

          Brillant

          • Re: (Score:3, Informative)

            The X4540 is even better to prove your point as it is cheaper than the X4500:
            12TB (48x250GB) - $21,995.00 [sun.com]
            and is virtually identical to the new 7210 box (config with 48x250GB) that sells for $34,995.00. Therefore proving that the same hardware is sold at a 60% markup ! Someone mod the parent up, he laid out the perfect counter-argument to the GP.
        • by segedunum (883035) on Monday November 10 2008, @07:43PM (#25714153) Homepage

          It's not like you can just grab 3 1TB SATA drives, throw them into RAID-5 and say that you've got 2TB of production ready storage. Well, you can, but you'd be an idiot.

          That's exactly what Google and many others do, and they spend their money, and significantly less than this, on managing that storage effectively. It works. When it boils down to it, you can have all the exorbitantly expensive and brilliant 'enterprise ready' tools you want but the bottom line is you need redundancy - and that's pretty much it.

          Your "home brew" solution will not meet any of the objectives Sun are achieving with this product.

          Sun say they are targeting small businesses, and they have lost already with this poor showing. They have advanced no further than when they stiffed all the Cobalt Cube customers and withdrew the product, who then went out and bought Windows SBS servers ;-). If you think people are going to jack them in for this then you need a stiff drink.

          Your spindle count will suck, so concurrent access will be slow.

          Ahhh, shit. I'm heart broken. What I'd like to know is how a small business will handle a behemoth like that, how they'll fund the electricity for all those drives and who'll manage it all. I expect that will be an ongoing cost to Sun support ;-).

          Keep to building crappy 3 or 4 disk RAID-5 systems using extremely large drives for storing your music, movies and pr0n on, but don't ever ever ever ever think about using those in any situation where your financial livelihood depends on that data.

          I have news for you. People have been doing it for years, and the reason why Sun's business has gone down the toilet to commodity servers, Linux and Windows, especially with small businesses, for the past ten years is exactly for this reason.

          Sun need to stop pretending that they can package up some commodity shit with some features very, very, very, very, very few need (and is waaaaaaaaaaaaay outside their target market) and label it as 'enterprise ready', which they think justifies an exorbitant price tag and ongoing support. They lost with this strategy with x86 and Solaris where they tried to protect SPARC, they lost with the exodus from SPARC after the dot com boom and they will keep on losing.

            • Re: (Score:3, Informative)

              For most companies, the "cheap crap" is more than adequate.
              They simply don't have the money to spend and they get by
              perfectly well despite not having drunk the cool-aid.

              Also, overpriced hardware with the label "enterprise"
              plastered on it is not going to do anything to prevent
              the need for a multiple data centers. Overpriced enterprise
              hardware and multiple data centers solve different problems.

          • by Hyppy (74366) on Monday November 10 2008, @10:39PM (#25715801)
            People avoid SATA in high-IO environments for a reason.
          • by Score Whore (32328) on Tuesday November 11 2008, @12:38AM (#25716617)

            Nice. How's the replication work on that rig you just built? And how many IOPS you getting? And how quickly will your vendor bring replacement hardware to you? How many filesystem snapshots can you take with your fancy ICH9 supporting linux? You gonna back that up over NDMP? How's the thin provisioning working out for you there? How much data you pushing through those two slots? Where's the other 2 gig ethernet ports. You got hot swappable power supplies there? After you're done stuffing all that gear onto your mobo, how many pcie slots you got left for future growth?

            No offense, but try and get some clue as to what it takes to have a commodity class storage appliance.

  • No RAID? (Score:5, Informative)

    by wonkavader (605434) on Monday November 10 2008, @05:44PM (#25712675)

    "All of the new unified storage systems include comprehensive data services at no extra cost, Fowler said. These include snapshots/cloning, restores, mirroring, RAID-5, RAID-6, replication, active-active clustering, compression, thin provisioning, CIFS (Common Internet File System), NFS (Network File System), iSCSI, HTTP/FTP and WebDAV (Web-based Distributed Authoring and Versioning)."

    Note that this system includes "RAID".

    • Re: (Score:3, Funny)

      by Anonymous Coward

      "All of the new unified storage systems include comprehensive data services at no extra cost, Fowler said. These include snapshots/cloning, restores, mirroring, RAID-5, RAID-6, replication, active-active clustering, compression, thin provisioning, CIFS (Common Internet File System), NFS (Network File System), iSCSI, HTTP/FTP and WebDAV (Web-based Distributed Authoring and Versioning)."

      Note that this system includes "RAID".

      (overheard in the Sun IT break room)

      "You know that fucking clueless Marketing guy? Yeah, he asked me to write up something for the new RAID-free array. Heh, I hooked him up."

  • by thanasakis (225405) on Monday November 10 2008, @05:47PM (#25712717)

    Considering that they've purchased MySQL, StorageTec and Cluster File Systems (of Lustre fame), developed ZFS, implemented CIFS in OpenSolaris from scratch (not Samba based), participated in NFSv4 and constructed the thumper, these machines hardly come as a surprise.

    For the last two years, almost all their moves are targeted towards one goal: Enter the storage market from a non-conventional angle. They want to do it unconventionally, because they know that storage more than anything else is becoming The commodity and today's toys won't cut it. Plus, at this point, all the mainstream storage vendors have difficulty tapping the low end. They may be able to sell their expensive products to clients with deep pockets, but for small businesses it's a different story. No to mention that they are unwilling to reinvent themselves. OTOH with all these inventions Sun may be trying to do what it did with workstations when it started in the 80s, start low and increase. Remains to be seen whether they can pull it.

  • by jcnnghm (538570) on Monday November 10 2008, @05:49PM (#25712745)

    This system will intelligently move the data around to put frequently accessed bits on the SSDs. This is a lot more than a 2u server with a few TB drives in a raid 10.

  • Zettabyte? (Score:3, Informative)

    by ethan0 (746390) on Monday November 10 2008, @05:52PM (#25712773)

    ZFS doesn't stand for zettabyte anything. "The name originally stood for "Zettabyte File System", but is now an orphan acronym." from wikipedia, sourced from http://blogs.sun.com/bonwick/entry/you_say_zeta_i_say [sun.com] .

    and of course "RAID Array" is lovelily redundant phrasing.

  • by More_Cowbell (957742) * on Monday November 10 2008, @05:59PM (#25712863) Journal
    People, please stop trying to compare a couple of drives from Newegg tossed in a chassis as a similar product for thousands less, simply because you have the same storage capacity.
    That's not even apples and oranges, it's more like apples and redwoods.
    Last I checked Netapp was still charging $10,000 per TB! [dedupecalc.com] Do you really think there is no reason for this?
    • by QuasiEvil (74356) on Monday November 10 2008, @07:04PM (#25713685)

      I hate to say it, but for the small business market, they should be compared. If you're selling a 2TB redundant storage device to a small business without a huge IT department, then you're competing against what can be built from commodity parts (aka, crap from Newegg + Linux + RAID) because often cost, not performance, is the defining factor.

  • RAID-Less how??? (Score:5, Insightful)

    by mkcmkc (197982) on Monday November 10 2008, @06:08PM (#25712969)
    • It doesn't use Redundant storage?
    • The storage isn't an Array? (Meaning what? That it's composed of non-uniform parts?)
    • It's not Inexpensive?
    • It's not Disk-based?

    The third one I believe--the rest I'm skeptical about...

  • oh ok... (Score:4, Interesting)

    by phaetonic (621542) on Monday November 10 2008, @07:57PM (#25714297)
    Fortune 500 companies typically standardize hardware, so people who say they can buy this from here, that from there, one more thing from eBay are rediculous.

    Also, to those who say small businesses can't afford this, its really an option. Some places like open source hodgepodges of hardware and some do not because their small business generates enough money that investing in enterprise class hardware with gold 4 hour response from a solid company with a history of UNIX experience and integration with Solaris.

    Also, said Fortune 500 companies get massive discounts, as what you're seeing is retail price.
  • So far the comments on this thread consist of "I could hack together some system for x% of the Sun price."

    The goal of this product is to compete with Netapp. If you've ever experienced Netapp licensing/pricing, this Sun solution is a bargain. People seem to be forgetting that this is a storage appliance.

    • Re:Looks great.. but (Score:5, Informative)

      by jcnnghm (538570) on Monday November 10 2008, @06:03PM (#25712919)

      Will that $600 box be using 14 146 GB 10k RPM SAS disks?

      These boxes aren't about providing stupid storage, their about providing massive I/O throughput. The larger boxes scale to 44TB and 576TB respectively. This also automatically moves frequently accessed data to flash drives (and RAM) for even faster I/O.

      These are absolutely monstrous compared to anything you could build for $600. There seems to be quite a bit of custom hardware to power this setup.

    • Re: (Score:3, Informative)

      With that said, linux REALLY needs ZFS , and not in userspace.

      Due to deliberate licensing issues we won't have native ZFS in Linux any time soon. However, BtrFS [kernel.org] should be merging into the mainline kernel soon enough (~2.6.29), and it includes most of ZFS's features plus a few of its own: storage pools, checksumming, mutable snapshots, built-in extent-level striping and mirroring, etc. It even supports in-place, reversible conversion from ext3 via a copy-on-write snapshot.

      • Re: (Score:3, Interesting)

        Due to deliberate licensing issues we won't have native ZFS in Linux any time soon.

        It's funny how this viewpoint is always the one promoted on slashdot. One could argue that the Linux GPL is the problem. FreeBSD and Mac OS X had no problem integrating ZFS into their code precisely because the ZFS license (CDDL) allowed it.

    • Re: (Score:3, Insightful)

      Yes yes yes, you can do that with just $1000 and a afternoon at Fry's or browsing Newegg, right?

      Everyone's missing the point here, and a lot of what is being said could be applied (just as wrongly) to NetApp... after all, those are just x86 boxes running a BSD kernel.

      The special sauce here is not so much the underlying OpenSolaris OS (which does provid the IO and services such as CIFS, NFS, iSCSI, data replication, and so on) but the Fishworks software put on top of it. Built-in failover clustering, the int

          • Re:DL180/185 (Score:5, Interesting)

            by level4 (1002199) on Monday November 10 2008, @10:38PM (#25715791)

            Guh. Sorry. I'm tired, and re-reading my comment the english is well-formed but the concepts are jumbled nonsense. Let me try again, by your leave...

            Yes, it's unavoidable to rebuild when you lose a disk, and there will be a performance hit unless you go for full on 100% redundancy, and not many companies can afford to do that with a lot of data.

            ZFS offers a number of benefits, though, in the event of drive failure-triggered rebuild, in that it basically knows where the data is and only bothers with that. A hardware controller has no idea what's data and what is blank space and so just redoes everything. In theory, assuming the MB/s of rebuild is the same, a ZFS rebuild of a half-full array should take half the time of a traditional controller.

            It is also much more intelligent about *what* it rebuilds, starting at the top and then descending down the FS tree, marking it as known good along the way. This means that if a second drive fails halfway through the resync, instead of a catastrophic failure you still have the data up to the point of failure.

            I can't remember where I read that; maybe here: http://blogs.sun.com/bonwick/entry/smokin_mirrors [sun.com]

            But I didn't even want to talk about drive-failure rebuilding, what I actually wanted to say that ZFS is, in theory, less likely to get itself into an inconsistent state in the case of power fluctuations, controller RAM failures, drive failures w/ pending writes, that kind of thing. That's the kind of rebuild I meant - after some kind of catastrophic failure. I should probably have said "integrity checking" though.

            By design, ZFS never holds critical data in memory only and so at least in theory should always be consistent on-disk. Basically it shouldn't need to fsck. That is a giant advantage to me, if it turns out to be as good in reality as it sounds on paper. Of course, that also has a lot to do with the capabilities of the FS proper, but removing the evil, evil HW controllers from the picture can only be a plus.

            I don't know why, but RAID controllers are the most unreliable pieces of hardware I have ever known, besides the drives themselves (but at least they are consistent and expected to fail). Get a few of them together and something WILL go wrong, more often than not in a horrible and unexpected way. When some RAM goes bad in a HW RAID controller you are in for a whole lot of subtle, silent-error-prone fun. Anything that gets the HW controllers out of the picture is a win for me.

            And don't even mention the batteries in HW raid controllers. They are the wrong solution to the power failure problem, especially since it's always after a failure that a disk will decide it's had enough of spinning and would just like to sit still for a while, thank you very much. Drive failure with pending writes! Exactly the words every administrator wants to hear. Almost as good as power failure with pending writes. Combine the two (highly likely!) for maximum LULZ. Ok, this is turning into a rant, I better stop.

            Anyway, thanks for the corrections. My original comment (and probably this one) came across as a confused mess upon re-reading .. sorry .. will sleep now : )