Apple shuts down ZFS open source project

24

Comments

  • Reply 21 of 72
    ouraganouragan Posts: 437member
    Quote:

    Apple's efforts to support the development of ZFS, an advanced file system originally created by Sun, were officially terminated today in a notice posted by MacOS Forge.



    The tersely worded message only stated that "The ZFS project has been discontinued. The mailing list and repository will also be removed shortly."





    ZFS seemed to be a good file system to adopt for future multi-core CPU computers. It didn't really matter if it wasn't fully backward compatible for older, less powerful, less advanced computers.



    But I do understand that patent suits can be an obstacle, especially if Apple cannot license the patents from NetApp, and I do understand that it can be problematic, less interesting to be the only one to develeop a file system if Oracle will not adopt or support ZFS for its own database computers.



    There remains BTRFS, the Oracle file system, which could possibly be extended with parts of ZFS, assuming that legal battles can be stopped. All should remember the fate of the late FireWire ports which were never widely adopted because royalties were too high. There is not much to be gained by sitting on technology.





  • Reply 22 of 72
    Quote:
    Originally Posted by mstone View Post


    Changing file systems is a daunting undertaking because the entire OS is affected in every possible way.



    Twenty years ago we had file system translators to help abstract the file system from the code. And that was on the Apple II.



    Obviously I can see the file system being an issue for boot code and disk utilities, but rest of the system really shouldn't care what the underlying file system looks like.



    At the end of the day, as long as the file system is secure, recoverable, and can survive a power loss or hard crash without corrupting, I don't care what it looks like either.
  • Reply 23 of 72
    sheffsheff Posts: 1,407member
    How sad, I heard a lot of good words about ZFS. Sucks that it did not work out.
  • Reply 24 of 72
    eluardeluard Posts: 319member
    Only a few days ago I was reading through the October archives of ZFS on MacOS Forge. This video was posted, which shows that there will soon be new features added to ZFS by its creators.



    https://slx.sun.com/1179275620



    There was then some discussion by one of the FS gurus (Dominic Giampaolo) at Apple as to whether ZFS could potentially lose data on a crash if the hard drive drops the flush

    track cache request. (He pointed out that all drives shipped by Apple since leopard do not do this ? and so would be safe. He was concerned about non-authorised HDs.) At around 23 minutes into the video there is a discussion of the proposed fix to this problem ? but the subsequent discussion was as to whether it was sub-optimal.



    I have no idea whether this issue was instrumental in this shutdown, and obviously Apple have not said anything with respect to reasons.



    But a sad day nonetheless for those of us who thought that a file system that enabled many HDs pooling together (plus snapshots!) was a good thing for the future of the Apple OS. I hope they choose a replacement candidate soon.
  • Reply 25 of 72
    bigpicsbigpics Posts: 1,397member
    Quote:
    Originally Posted by wizard69 View Post


    The other thing here is that ZFS isn't exactly the simplest thing going. If Apple is looking for something new they need to look at the open source and Super Computing worlds. Linux has a number of interesting file systems and there are a couple associated with super computing.



    Something, it seems, will have to give if Apple OS's are to reach their fuller potential. So.... ...OS X is *nix and includes - and contributes to - a lot of major open source code... ...so what do the "nixes use and what are the strengths and weaknesses?



    'From scratch,' with Apple's current chops, could be cool, but why re-invent something worthwhile that can be polished in shiny Apple goodness and re-released as open source to get wide-spread adoption as a standard that would ultimately benefit them. Just as releasing GCD under an Apache open-source license is likely to do.



    If a good FS candidate exists, of course.



    Supplanting the Redmond hegemonists - if they're ever displaced - will be a long twilight war fought on many fronts. Apple has broken away in hand-held computing for example, which will be at least as big as PC's have been or much bigger. With MS in a distant 4th or 5th place or so (Blackberry, Android, Apple, Nokia).



    But PC and server computing are hardly in the dust bin of digital history. And here Apple has two fronts of attack that no other company has - they can function simultaneously as the most proprietary computing company most in charge of what it designs, makes, markets and services AND as a major creator of the ever-more important open source code that permeates the internet, the enterprise - every market actually since every connected computer is a client computer of some other computer more or less.



    The more deep code Apple releases and sees widely adopted (while keeping its UI secret sauces its own) the closer we are to the day when Apple OS's are more built on world standards (influenced and updated by Apple) than MS's. It's all about outflanking and surrounding the other guys and by allying itself with Open Source, the more Apple has as many allies and troops with a common interest as MS does with its box makers, etc.



    Also you'll notice the word "Wintel" has pretty much dropped off the planet since Intel's affair with Apple. Another underappreciated strategic move (I really think they "think big" and long-term out in Cupertino) with implications beyond salvaging their last shot at being relevant in hardware.



    PS: Bashing Flash while encouraging tools to be built around HTML 5 is another smart move in the long run - that only Apple had enough panache to pull off on its own without being punished by the market (tho' it also hinged on getting YouTube to recode everything into H.264). By all reports all the iPhone killers that have Flash on their features checklist perform abysmally.



    PPS: I love that giving really cool things away turns out to make viable market sense in the digital age. Go open source!
  • Reply 26 of 72
    solipsismsolipsism Posts: 25,726member
    Quote:
    Originally Posted by bigpics View Post


    PS: Bashing Flash while encouraging tools to be built around HTML 5 is another smart move in the long run - that only Apple had enough panache to pull off on its own without being punished by the market (tho' it also hinged on getting YouTube to recode everything into H.264). By all reports all the iPhone killers that have Flash on their features checklist perform abysmally.



    If your goal is to stream video then Flash is a bad choice for phones, but I have to say that many of these HTML5 advances, like Canvas, are really processor intensive. For plain interaction of a website, Flash is currently easier to code for while offering more functionality. That said, I welcome the demise of Flash’s ubiquity. Those sites are just annoying. I don’t need splash pages!
  • Reply 27 of 72
    This is a real tragedy. Realistic decision, but a tragedy none the less.



    ZFS is the easiest way to ensure that you don't have data corruption for large data stores, especially at a consumer level! I hope Apple gets some kind of checksum based, copy-on-write file system, even if it isn't ZFS. I'd love to see a Drobo that supports it... but that requires OSX file system support.
  • Reply 28 of 72
    This IP law and lawsuit issue is starting to really get in the way of progression and advancement of mankind.



    Xserve Raid. Some of us called it Xraid implying Xserve-Raid storage array...

    http://www.apple.com/server/storage/



    When Sun was looking for a new owner I thought for sure Apple would jump.

    Pros:

    1) Right down the street (sorta).

    2) Large office building and massive talent pool.

    3) Good server client base.

    4) Interesting technologies both software and hardware that could use a dash of Apple.

    5) License of many patents.

    6) Easy porting of current Xcode systems to Solaris to bring more x-platform development and enable current Obj-C developers a second platform and allow easy transfer to OS X.



    The list goes on and on. Cons are there but we are talking about a company that went to the brink and came back as a leader. They could fix it and get everything they wanted out of Sun. There's a lot there at Sun and Oracle's bid was a bargain. I know some of the Sun shareholders would have gladly traded their shares for Apple stock saving Apple from dipping into their pockets and most Sun folks may actually like having their friends helping them improve.



    Apple is and thankfully leading down a path of solidarity in the industry relying less on outside corporate help and using more open technologies. In the process then end up being more "standard" than others and allow a free-flow platform for future development in the computer structure itself.
  • Reply 29 of 72
    I was also surprised that Apple didn't jump at the chance to get Sun. So many sinergies to be exploited, plus Apple having a great back door to the Enterprise market. Who was sleeping on this one?
  • Reply 30 of 72
    Quote:
    Originally Posted by shadow View Post


    Apple did a lot of FS related changes in Snow Leopard. One of the [several] reasons behind dropping 64-bit Carbon could be the future change of the FS in mind. In Cocoa, the internal implementation of the filesystem is well abstracted and Cocoa applications will not be affected. Only a handful of UNIX tools need to be adapted to the new FS (UNIX works on different filesystems anyway).



    The next OS release could drop Carbon support altogether. This will clean the way for a new FS. If the rumored Marble UI is added, that would make Mac OS 10.7 a major overhaul.



    That said, changing the default file system by all accounts is not an easy step. But it is doable, much more on Mac OS than on Windows.



    This brings another point to the Windows vs Mac OS debate:

    No matter how good Windows 7 is, Windows is heavily anchored by legacy code. The progress on the Windows side is going to be about the same as it was during the last decade (9 years to get a stripped-down Longhorn to be finished). The last anchor for Mac OS is Carbon. When Apple breaks the chain the already decent pace of Mac OS development will accelerate even further!



    I?m not sure what makes that more wrong:



    1. The fact that Carbon works on other file systems just as well as Cocoa does (and did work, in fact, on the experimental read/write ZFS drivers that were available on MacOSForge until just recently),



    2. The fact that the bits of Carbon that interface with the file system weren?t even removed from 64-bit, just the UI parts, or



    3. The fact that Cocoa uses Carbon internally all over the place for file system stuff.
  • Reply 31 of 72
    mr. hmr. h Posts: 4,870member
    Quote:
    Originally Posted by John the Geek View Post


    No, he meant XServe RAID. I should know, I have one. It's not called an XRAID.



    Exactly. What's needed here is the wayback machine to prove it:



    http://web.archive.org/web/200407050...m/xserve/raid/
  • Reply 32 of 72
    Quote:
    Originally Posted by bloggerblog View Post


    No such thing, it's called XRAID.

    This is typical Apple, and that is exactly the kind of half-baked promises that keeps IT people far away from Apple.



    Xserve RAID is the name of the Apple legacy product.



    XRAID is a product from Active Storage http://www.getactivestorage.com/overview.php



    The Apple Xserve RAID was long over due for a upgrade. Is it very slow compared to the Promise RAID that Apple sell instead these days. The Xserve RAID did not have redundant controllers (very critcal) and gets very slow when under 200 GB of space left. The Promise VTrak does not have these shortcommings and that at the same price. The Xserve RAID had ATA disks that now is becoming almost impossible to get hold of in larger than 500GB sizes.



    Apple should continue to do what they do best. Leave enterprise storage to companies that specializes in this. The actual storage manufacturer does not matter me.
  • Reply 33 of 72
    blastdoorblastdoor Posts: 3,255member
    Quote:
    Originally Posted by DaveGee View Post


    Yea x12 for me... well not me ..personally.. but the organization that I worked for in a previous life, we had a dozen of em.



    'No such thing, it's called XRAID.'



    Nothing I hate more than people who come off like the book of knowledge and spew nonsense that impressionable readers mistakenly take for facts.



    Agreed!!
  • Reply 34 of 72
    blastdoorblastdoor Posts: 3,255member
    Quote:
    Originally Posted by ThePixelDoc View Post


    I was also surprised that Apple didn't jump at the chance to get Sun. So many sinergies to be exploited, plus Apple having a great back door to the Enterprise market. Who was sleeping on this one?



    I was not at all surprised. Apple has never bought a company that large, and I don't blame them. Integrating two large companies is a major effort that requires a great deal of senior management's time. If senior management is spending time on that, then they're spending less time on paying attention to products. Look at MS -- Ballmer likes to think about big blockbuster mergers and other MBA-ish stuff instead of products. He delegates all the product stuff to people further down the pike, and what you get is a great deal of variation in product quality with little strategic coordination/direction.
  • Reply 35 of 72
    Quote:
    Originally Posted by ThePixelDoc View Post


    I was also surprised that Apple didn't jump at the chance to get Sun. So many sinergies to be exploited, plus Apple having a great back door to the Enterprise market. Who was sleeping on this one?



    I believe that the mind set at Apple is to buy a specific technology & talent pool, rather than buy a diverse company with baggage they don't want or need.



    If they can't easily buy the specific parts that they want, Apple will usually hire a few key people and build it (whatever) in-house.



    *
  • Reply 36 of 72
    vineavinea Posts: 5,585member
    BTRFS is a non-starter given it is GPL unless Apple licenses BTRFS from Oracle. IF Oracle even has all IP rights to it.



    Dropping ZFS is a shame but it's probably from the patent issues and Oracle not going to bother defending ZFS from NetApp.
  • Reply 37 of 72
    Quote:
    Originally Posted by maercsrats View Post


    ZFS isn't dead.

    ...



    My bet is that apple will push form something like the HAMMER FS (http://en.wikipedia.org/wiki/HAMMER). It's got a BSD license and some of the same goals of ZFS.



    It's a shame about dropping ZFS, because it has some really useful features - not least of which is the goal of data integrity (HAMMER, so far, only seems to mention CRCs on metadata.) Is anyone continuing, independent of Apple? The MacFUSE plugin had support, I think?



    Only the other month, when comparing identical filetrees of photographs on two different HFS+ disks, I had one RAW photo file show up with 512kB of nulls in the middle.



    Even with single-disk systems, ZFS allows for storing two copies of data on a drive, to insure against block corruption (better than corrupt data, and corrupt backups...)
  • Reply 38 of 72
    The patent issue(s) would seem to be a non-issue according to the terms of the CDDL which ZFS is licensed under:



    Quote:

    What does the CDDL say about patents?



    The CDDL provides an explicit patent license for code released under the license. This means that you can use, modify, and redistribute code released under CDDL without worrying about any patents that the contributors of the code (including Sun) might have on the contributed technology. The license also includes a provision to discourage patent litigation against developers by revoking the rights to the code for anyone initiating a patent claim against a developer regarding code they have contributed.



    So, as far as Apple is concerned, NetApp wouldn't have any judicial standing suing Apple for using ZFS, they'd have to go after Sun or Oracle (once the acquisition is complete). Yes, any court decisions could affect Apple downstream in the long run, but this could be the case for any GPL, BSD, or other Open Source software or libraries used by Apple and others. It's a known and accepted risk of doing business in the software world today and is the primary reason why software patents need to be voided in general. This is why Linux supports so many filesystems out of the box - there is always a fallback with a relatively painless migration path in case of an issue - legal or technical-wise.



    Apple shouldn't be so hast in their decision, ZFS is still a great file system especially from a storage perspective with its inherit robustness and disaster recovery capabilities. NTFS and HFS(+) do not nearly compare with ZFS's feature set.



    However, I don't understand why Apple doesn't just adopt ext2/3 like just about every other Linux system out there or UFS1/2 like the BSDs if they want to be more like them architecturally-wise. Of course it really wouldn't matter what they used if they adopted Fuse in the first place.
  • Reply 39 of 72
    irelandireland Posts: 17,798member
    They also cancelled the "make-a-real-mouse" project.



    This is a pity, ZFS looked promising for Apple. It's not the worst thing Apple has done lately though, not adding an SSD option for the new iMac took the biscuit. People who defend Apple on this need to seriously take a loot at their fanboy status. The word is: OPTIONAL!!!



    It's not exactly specific to this thread, but now I shall take a look at adding an SSD to my 24" iMac. I'm doing it this time for definite.
  • Reply 40 of 72
    seek3rseek3r Posts: 179member
    Quote:
    Originally Posted by bloggerblog View Post


    No such thing, it's called XRAID.

    This is typical Apple, and that is exactly the kind of half-baked promises that keeps IT people far away from Apple.



    No. Answers for what keep IT people away from apple could include but are not limited to:



    Problematic NFS implementation



    High cost for compute nodes make clustering impractical



    Lack of consistency in usage of config files



    Lack of low-cost entry to the server side of things (this just changed, so we'll see how it effects things)

    ---

    All of those could have been trotted out as arguments against apple's adoption in the machine room without argument from me.



    A lack of a disc array in the product line up and the use of HFS+ as a file system really aren't generally why Apple isnt chosen.
Sign In or Register to comment.