Forgot your password?
typodupeerror
Microsoft Software Windows Technology

Microsoft Announces ReFS, a New Filesystem For Windows 8 459

Posted by timothy
from the comes-with-striped-shirt dept.
bonch writes "Microsoft has shared details about its new filesystem called ReFS, which stands for Resilient File System. Codenamed 'Protogon,' ReFS will first appear as the storage system for Windows Server and later be offered to Windows clients. Microsoft plans to deprecate lesser-used NTFS features while maintaining 'a high degree of compatibility' for most uses. NTFS has been criticized in the past for its inelegant architecture."
This discussion has been archived. No new comments can be posted.

Microsoft Announces ReFS, a New Filesystem For Windows 8

Comments Filter:
  • by TechGuys (2554082) on Tuesday January 17, 2012 @11:23AM (#38724884)
    After my initial tests, I must say that ReFS is incredible advangement. ReFS supports named streams, object IDs, short names, compression, file level encryption (EFS), user data transactions, sparse, hard-links, extended attributes and quotas. It is basically all the best filesystems compiled into one.

    Not only is this good for Windows system, but overall network architecture.
    • Re: (Score:3, Interesting)

      by Anonymous Coward

      How does it compare to ZFS in terms of resilience? After all, it's in the damn name.

      • by Creepy (93888) on Tuesday January 17, 2012 @03:07PM (#38727974) Journal

        The "resilience" is from copy on write (CoW), which is used in Volume Shadow Copy and Microsoft SQL server. It is also able to cloud data across multiple volumes on different machines from what I read. Since both CoW and ZFS's copy work a lot like RAID0 (as far as I can tell), I'd expect them to be similar in this respect, however ZFS also does checksum tests and NTFS doesn't BUT I don't know if ReFS will or not.

        That said, ZFS is a WAY better file system, and I'll give you a few reasons why:
        No max path length restriction (TFA says there still be one for ReFS)
        Variable Block sizes and Sparse Files
        Allocate on Flush [wikipedia.org]
        Block Journaling (aka Journaling File System) as opposed to Metadata only Journaling (NTFS and probably ReFS) which is less reliable
        Logical Volume Management [wikipedia.org]
        and that is just naming a few off the top of my head with some links to what they mean if it seemed like it may not be obvious (the others are fairly commonly talked about IMO - if you don't know them, they should be easy to search for)

          I'm fairly certain NTFS still doesn't support user metadata, either, and I believe zfs does (most modern FS's do), so I doubt ReFS will (what I mean by this is I can tag a piece of data as, say "photos" and then when I search for photos, those are found first - this is a feature like what was planned for WinFS's and what Apple's Spotlight does).

    • by Anon-Admin (443764) on Tuesday January 17, 2012 @11:26AM (#38724946) Homepage Journal

      So they are starting to catch up with the ext3 filesystem.

      • Re: (Score:3, Informative)

        by vlm (69642)

        So they are starting to catch up with the ext3 filesystem.

        I thought it sounded pretty much like a dumbed down version of AFS... from the early 90s. The problem is I never use any of that extra stuff because I have no use for it. I don't remember if I can do sparse with AFS because I don't care about sparse. At home I do the openafs thing for linux, mac, and windoze and everything is in AFS, so I don't really care what windows uses natively, its just kind of a bootloader to get to my real files over afs.

        I hope there is a way to disable file level compression, be

      • Supposedly they also support pools across multiple devices of different sizes that can be reallocated dynamically.

        In that regard, it is more like zfs and btrfs, and on par with the best filesystems out there.

        I'm curious what performance is like.

    • by gnalre (323830) on Tuesday January 17, 2012 @11:27AM (#38724948)

      After my initial tests, I must say that ReFS is incredible advangement. ReFS supports named streams, object IDs, short names, compression, file level encryption (EFS), user data transactions, sparse, hard-links, extended attributes and quotas. It is basically all the best filesystems compiled into one.

      Not only is this good for Windows system, but overall network architecture.

      and of course will be an open standard(Sarcasm Alert)

      • by hedwards (940851)

        That thought crossed my mind, there are legitimate reasons for ditching NTFS, but I can't help but wonder if this has anything to do with Win 8 being the first release in a long time where they weren't under DoJ supervision. Also, at this point, NTFS support on other OSes is pretty good. Seems like a really convenient way of making it inconvenient to interoperate or multiboot.

    • by DarkOx (621550) on Tuesday January 17, 2012 @11:28AM (#38724958) Journal

      That's very interesting given the article says

      There are some NTFS features for which Microsoft plans to drop support with ReFS, specifically named streams, object IDs, short names, compression, file level encryption (EFS), user data transactions, sparse, hard-links, extended attributes, and quotas, Verma blogged. That said, one of Microsoftâ(TM)s goals with ReFS is to âoemaintain a high degree of compatibility with a subset of NTFS features that are widely adopted while deprecating others that provide limited value at the cost of system complexity and footprint,â Verma said.

      • by mlts (1038732) *

        Some of those features are actually useful:

        Compression comes into handy for dealing with directories full of log files.

        File level encryption is useful for volumes where BitLocker can't be used.

        Sparse files are extremely useful.

        As for quotas, unless they have another layer for warning/enforcement, how will places keep users from filling up their home directories?

        I'm hoping ReFS is up to ZFS with needed features, such as deduplication, encryption, an analog of RAID-Z, the filesystem working with the LVM layer

    • by ledow (319597) on Tuesday January 17, 2012 @11:28AM (#38724960) Homepage

      Great. What does that mean for users?

      Amateurs: Nothing at all.
      "Know-a-bit's": Almost nothing.
      Professional users: Nothing we couldn't do before.
      State-of-the-art, top-dog, storage-gods: Nothing very special or new at all.

      Now, if you'd said that it finally supported WinFS-style file tagging and searching, then you'd have ticked lots of boxes for all manner of users. As it is, it's a "slightly better filesystem than before" and hardly newsworthy (out of all your "features", I only spot one that you can't already do with Windows alone and that would ever be exposed to someone NOT using bit-level access to the drive - file level encryption).

      • by vlm (69642) on Tuesday January 17, 2012 @11:32AM (#38725036)

        End users: We still live in a world of 8.3 filenames. Sorry. Till the last PC is burned in a bonfire...

        You know what would be a funny graph of google data? How many are still serving up .htm files instead of .html files vs year.

        20 years from now my grandkids are going to have to answer on Jeopardy why computer filenames are still in a 8.3 filename format.

        • by peragrin (659227)

          No we will live in 8.3 file names as long as FAT is the defacto standard for all portable drives.

          My only wish is for MSFT to support a filesystem they didnt create. That way a third paty FS can get used for portable media

          • by exomondo (1725132)

            No we will live in 8.3 file names as long as FAT is the defacto standard for all portable drives.

            Why? FAT LFN support has been commonplace for well over a decade. Where have you recently been restricted to 8.3 filenames?

    • Umm, aren't those the features being dropped?
    • by eldavojohn (898314) * <[moc.liamg] [ta] [nhojovadle]> on Tuesday January 17, 2012 @11:32AM (#38725032) Journal

      After my initial tests

      Wait, what? From the article:

      Officially named ReFS — for Resilient File System — the new file system will be made available via a staged “evolution,” according to a January 16 post on the “Building Windows 8 blog.

      So you're saying something that was just announced and will be made available via a staged evolution has already been tested by you? Impressive!

      It is basically all the best filesystems compiled into one.

      Thanks for summing it up for me there, bud. I didn't realize it was the greatest goddamn filesystem I could imagine, why didn't you just say "Imagine what your dream filesystem will be able to do, this is it." I wonder though, will it have the homicide capacity of ReiserFS?

      This reminds me of my initial tests of cold fusion. I must say that cold fusion is incredible dvangement. Cold fusion supports providing us with unlimited power from a glass of water, it prints money, it gives the user eternal life, it allows the user to travel faster than the speed of life and -- when activated -- attractive women jump out of the core reactor demanding money shot after money shot.

      • by anonymov (1768712)

        It gets better:

        TFA:

        some NTFS features for which Microsoft plans to drop support with ReFS, specifically named streams, object IDs, short names, compression, file level encryption (EFS), user data transactions, sparse, hard-links, extended attributes, and quotas, Verma blogged.

        His post:

        I must say that ReFS is incredible advangement. ReFS supports named streams, object IDs, short names, compression, file level encryption (EFS), user data transactions, sparse, hard-links, extended attributes and quotas

        He's either overpaid as MS shill, or underpaid as dark PR style "obviously dumb MS shill" troll.

        • by cbhacking (979169)

          Or... he's being sarcastic, and you are a fool. That list was basically a complete rundown of NTFS's newest or least-known features (some, like compression and EFS, have existed for over a decade). It was quite obvious, even without reading TFA, that some if not all of them were going to be cut from the new version.

          Mind you, if you want me to use a filesystem that doesn't support transactions, it better have some other damn good perk. Compare what happens if a big file operation (say, copying a large direct

    • by andydread (758754)
      Yes troll but it features POOR interoperability. No thanks.
    • After my initial tests, I must say that ReFS is incredible advangement. ReFS supports named streams, object IDs, short names, compression, file level encryption (EFS), user data transactions, sparse, hard-links, extended attributes and quotas. It is basically all the best filesystems compiled into one.

      We have a cautionary fairy tale here about the cat and the dog baking a cake. Since they don't know how to bake a tasty cake, they decide to put every tasty ingredient they know into the mix. So then naturally include flour, milk, butter, eggs, whipped cream, bones, mice, sausages, candies, chocolate, pork cracklings, sauce...and so on, except for bread, which neither cats nor dogs seem to like. Well, I guess you can guess what they got in the end.

  • by uigrad_2000 (398500) on Tuesday January 17, 2012 @11:27AM (#38724954) Homepage Journal

    This is a bad idea.

    Now we can count on some guy named 'Hans Resilient" to be tried and found guilty of murder.

  • Interesting (Score:5, Interesting)

    by msobkow (48369) on Tuesday January 17, 2012 @11:29AM (#38724998) Homepage Journal

    I can't say that I've ever used any of the NTFS features they're planning to drop.

    I do wish Windows had a sane soft-link system like *nix does; I've yet to run into an application that automatically dereferences a .lnk when opening it. You have to futz around with opening the link manually, reading it's redirect, and then opening THAT instead. Very crude and ugly.

    But more to the point, I didn't see much about what might be NEW with this file system, only what's OLD and being discarded.

    Mind you, some basic feature cleanup never hurt anyone. But if that's the case, why not NTFS2 instead of a marketing buzzword?

    • by msobkow (48369)

      Typo: I meant "API", not "application".

    • Re:Interesting (Score:5, Informative)

      by EvanED (569694) <evaned@gma[ ]com ['il.' in gap]> on Tuesday January 17, 2012 @11:35AM (#38725086)

      I do wish Windows had a sane soft-link system like *nix does; I've yet to run into an application that automatically dereferences a .lnk when opening it. You have to futz around with opening the link manually, reading it's redirect, and then opening THAT instead. Very crude and ugly.

      Man, if only [microsoft.com].

      (OK, it's not quite sane considering you have to distinguish between links to files and links to directories at creation time. I'm not sure what happens if you flip it behind its back.)

    • by Malc (1751)

      Why are you referring to shortcut files? That's something entirely different.

    • Re:Interesting (Score:5, Informative)

      by anonymov (1768712) on Tuesday January 17, 2012 @11:48AM (#38725280)

      There's a blog post [msdn.com] linked from the article.

      There's all kinds of promising stuff, like data corruption resilience and dropped/extended limits.

      Much more interesting read than the linked ZDNet article.

    • Re:Interesting (Score:5, Informative)

      by SpryGuy (206254) on Tuesday January 17, 2012 @11:51AM (#38725332)

      But more to the point, I didn't see much about what might be NEW with this file system, only what's OLD and being discarded.

      Let's see: 32K file name and path limits (instad of 255), on-line recovery from corruption (no more "Check Disk" or offline recovery-rebuild), faster performance, built in recovery of data on failed disks (via Storage Spaces), hot-adding-more-storage to volumes, better control of allocation and localization on the drive, attribute checksums (and auto detection and recovery from "bitrot")....

      Did you RTFA at all?

      • ... to no one. Apart from maybe malware writers who'll be able to put an entire virus in the filename. Whether they'll be able to hide it or even use it is another matter but I wouldn't put it past Windows to have a nice exploit available.

        • by Thing 1 (178996) on Tuesday January 17, 2012 @01:25PM (#38726650) Journal

          ... to not one.

          The real world disagrees with your statement: we have TFS projects with long directory and file names, such that we cannot map the entire TFS source in a single folder. Even naming it e.g. "c:\x" (or "d:\", putting it on a separate drive), the paths and files still exceed MAX_PATH (which is 260, [microsoft.com] not 255).

          So, this feature will be useful to our shop.

          It's also useful for "rolling backups"; I administer family machines, and one has been upgraded from a desktop, to a laptop, to another laptop. The first upgrade, I copied all the files to "c:\e" (old machine was an eMachine). That laptop died, we used a restoration company that started with a "G" to get the data back (now we backup via WHS), and I saved that in "c:\g" (so there's a "c:\g\e" with the desktop's files). The third machine (second laptop) has "c:\h" (which also contains "c:\h\g\e"). Other times I've saved backups with more descriptive names, like "Backup of the Dell Inspiron 5150, 2011-11-11", and sometimes those backups fit inside each other like expressed above.

          So, I have examples from both home and work where having longer-than-MAX_PATH file/path names would be useful.

    • by DarkOx (621550)

      NTFS supports softlinks, (junction points) its just none of the user land stuff that ships on the Windows platform knows how to deal with it.

      Explorer for instance can't create them, and indicate that something is a link, and can't correctly total up disk usage for a tree if you have used them in that tree.

       

  • Great! (Score:3, Insightful)

    by Sez Zero (586611) on Tuesday January 17, 2012 @11:29AM (#38725000) Journal
    All we need is another MS-specific filesystem to cause compatibility headaches.
    • Re:Great! (Score:4, Insightful)

      by uigrad_2000 (398500) on Tuesday January 17, 2012 @11:51AM (#38725340) Homepage Journal

      I'll agree.

      As ugly as NTFS is, the one thing I've liked about it is that it's the only FS used by Windows and Windows Servers for a dozen years.

      With Linux, on the other hand, I've had to deal with ext2, ResierFS, ext3, ext4, and those are only the popular ones! There are a ton of other specialized filesystems for other features, such as encryption or use on flash memory!

  • by snsh (968808) on Tuesday January 17, 2012 @11:34AM (#38725064)

    Dropping support for compressed folders and hard links? I use those features all the time. Especially when you troubleshoot a server with a subfolder containing 12GB of log files, and have no direction or policy about what to do with those old log files, you could safely enable compression on the folder and they magically take up less space.

    • by MagicM (85041)

      compressed folders + truecrypt + robocopy also makes a wonderful hassle-free backup system.

  • by Robert Zenz (1680268) on Tuesday January 17, 2012 @11:34AM (#38725066) Homepage
    From the blog post: [msdn.com]

    Today, NTFS is the most widely used, advanced, and feature rich file system in broad use.

    If this is true...it's a very sad world we live in...

    • by Sycraft-fu (314770) on Tuesday January 17, 2012 @12:33PM (#38725900)

      When it comes down to it, NTFS is a pretty good file system. If you look in to things you find that the feature list for BTRFS reads an awful like a feature set of current NTFS.

      None of that is to say that NTFS couldn't stand improvement, and indeed it is being improved, but I've yet to see the amazing widely used file system that is so much better than it. Ext3 is functional, but leaves much to be desired.

  • 1) which problem does this solve ?

    2) if the answer at #1) is not "null", then how monkeyproof is it ?

  • Warning (Score:5, Funny)

    by TBedsaul (95979) on Tuesday January 17, 2012 @11:38AM (#38725138)

    If you're married to "Hans Resilient", you'll want to start running now.

  • by Sir_Sri (199544) on Tuesday January 17, 2012 @11:40AM (#38725162)

    Sounds like they're due for a refresh so they can get some new patents on their filesystem to make sure all the device makers need to continue to pay them money.

    • Sounds like they're due for a refresh so they can get some new patents on their filesystem to make sure all the device makers need to continue to pay them money.

      That's what exFAT was for. This is probably to make it harder for non-Windows folks again since they can finally read/write to NTFS (stable since late 2.6.2x series for Linux; don't know about the others).

  • by Moses48 (1849872) on Tuesday January 17, 2012 @11:50AM (#38725326)

    I'm not a filesystem guru. I stick to programming in the application space mostly. But I have noticed a large time discrepency compiling a large project using EXT4 vs NTFS. EXT4 being multiple times faster then doing the same compile on an NTFS. My question now is, will ReFS bring those times up to similar values?

    PS. Also looking at the dropped support for short names, i think quite a few server batch files will be broken.

    • by SpryGuy (206254)

      I've heard anecdotal evidence (so take with a grain of salt) that doing stuff on ReFS is much faster.

      Keep in mind this initial release is for servers only, and NOT for boot volumes, so it'll be a while (half a decade or more) before it trickles down into most desktops/laptops.

      • So this is for servers, not for boot devices

        So what is it for, for storage most people are using network storage, where the filesystem is very often not Windows Native?

  • by IGnatius T Foobar (4328) on Tuesday January 17, 2012 @11:57AM (#38725436) Homepage Journal
    "Microsoft plans to deprecate lesser-used features" --- such as the reasonable level of compatibility that has started to show up in non-Microsoft implementations of NTFS over the last couple of years. We may be assured that ReFS is a patent minefield.
  • by QuietLagoon (813062) on Tuesday January 17, 2012 @12:06PM (#38725552)
    ... it was hyped, among other things, as a file system that would never need to be defragmented.

    .
    I have to wonder how much of the pre-release ReFS hype will prove to be true in the coming years.

    • And they specifically used the same FS number as HPFS which Microsoft and IBM jointly developed for OS/2, and was also used by Apple for MacOS specifically to keep systems from being compatible.

      So, watch them use partition table FS identifier 82 (Linux FS - Ext2/3/4) instead...for the same reason. However, that has typically only hurt Microsoft as everyone else figures out how to be detect the differents FS's on the actual partition, especially the Linux folks.
      • by drsmithy (35869)

        And they specifically used the same FS number as HPFS which Microsoft and IBM jointly developed for OS/2 [...]

        The reason for that should be pretty obvious - NTFS was supposed to replace HPFS in what was then still called OS/2 NT.

        NTFS development started along with NT development in the late '80s. Back when IBM and Microsoft were still BFFs and *years* before they would have any thoughts about "compatibility".

  • by BoRegardless (721219) on Tuesday January 17, 2012 @12:09PM (#38725592)

    All the file utilities for both Mac and PC and how you handle these different systems including forward/backward compatibility, Parallels, VMWare, Backup software, hard drives and tape devices will all go through teeth nashing debugs as we try to get everything to work with a new file system.

    That may be OK when you are an IT professional.

    For someone who "just wants it to work" there is likely to be lots of surprises ahead.

  • NTFS is resilient! (Score:5, Interesting)

    by DigiShaman (671371) on Tuesday January 17, 2012 @12:16PM (#38725690) Homepage

    A few weeks ago, I pulled "Hail Mary" with regards to saving an SBS 2003 server. For whatever reason, the server would not boot after a power failure. The RAID cache was not dirty on the card, and the RAID volume passed a manual parity consistency check. Unfortunately, the server would still not boot into the OS. It kept throwing a BSOD or hung at finding the hal.dll file. Attempting to access the recovery console or other F8 invoked options failed. Any Server 2003 disk would throw a BSOD the moment it attempted to mount the boot "C" volume. It wasn't the RAID drivers, but actual NTFS corruption causing the kernel panic. Serious shit. However, a Server 2008 R2 disk did save my ass. I was able to mount the volume through a command recovery console. A chkdsk revealed massive amounts of corruption. Server is fucked right? NO! A "chkdsk /R" command was able to find and repair all errors. No data loss what-so-ever.

    Basically, the server must have been busy with installing updates or something when the power died. An old UPS battery will do that. But this goes to show how remarkably resilient the NTFS system is. Absolute respect!

    • by Anonymous Coward on Tuesday January 17, 2012 @01:05PM (#38726362)

      Respect?? That's absolutely terrible.

      A modern journalling filesystem should not experience any corruption after a crash, because journal recovery is supposed to keep data structures consistent.

      Not only that, but NO filesystem, journalling or not, should cause a kernel crash if it is corrupted.

      Microsoft has done one thing well, and that is to lower the expectations of their users so far, that what should have been a few second journal recovery turned into a big outage and manual recovery of a massively corrupted filesystem, and that gains them "Absolute respect".

  • by Fujisawa Sensei (207127) on Tuesday January 17, 2012 @12:35PM (#38725940) Journal
    With patented algorithms to ensure that other OSes won't inter-operate without paying the Microsoft Patent Trolls a fee high enough to buy a Windows license for each machine in question.
  • by Anonymous Coward on Tuesday January 17, 2012 @01:18PM (#38726540)

    It needs some way to securely mount a remote filesystem. SMB and non-anonymous FTP shouldn't be used over the internet ever. It wouldn't be too bad except that FTP is incredibly difficult to reliably tunnel due to it opening connections in both directions on random ports. I would be a happy person if Windows added native support for sftp.

"What the scientists have in their briefcases is terrifying." -- Nikita Khrushchev

Working...