Native ZFS Is Coming To Linux Next Month 273
An anonymous reader writes "Phoronix is reporting that an Indian technology company has been porting the ZFS filesystem to Linux and will be releasing it next month as a native kernel module without a dependence on FUSE. 'In terms of how native ZFS for Linux is being handled by this Indian company, they are releasing their ported ZFS code under the Common Development & Distribution License and will not be attempting to go for mainline integration. Instead, this company will just be releasing their CDDL source-code as a build-able kernel module for users and ensuring it does not use any GPL-only symbols where there would be license conflicts. KQ Infotech also seems confident that Oracle will not attempt to take any legal action against them for this work.'"
Good Article (Score:5, Insightful)
No, really. I had a bunch of questions going in, and they were all answered. This is rare enough to warrant a shout out to Michael Larabel.
I disagree with some of his subjective claims like x86_64 being a substantive limitation or ZFS on Linux remaining niche (I guess that depends on how you define the niche...) but he got the national lab project, the zpool version, the Oracle (nee Sun) patent problem. Kudos.
FreeBSD 9 is probably where ZFS will wind up finding a proper home, I'm guessing.
Re:Good Article (Score:5, Insightful)
How do you think it is not a substantive limitation?
My phone runs linux and is not x86 of any shape or register size, nor is my workstation, nor are many other machines I have running linux. This is just like people who think flash working only on x86 32bit linux is good enough.
If FreeBSD ever gets a good ZFS implementation expect lawsuits to commence.
Re: (Score:2)
Sadly, yes, there is every reason to expect it.
Re: (Score:2)
If FreeBSD ever gets a good ZFS implementation expect lawsuits to commence.
None have been filed since it was production-ready last year. Besides, what would they sue over? The FreeBSD team using code that Sun deliberately and explicitly licensed for such things?
Re: (Score:3, Informative)
None have been filed since it was production-ready last year.
It's not. Yet. There are many reports of lock-ups with uptimes on the order of a week. Soon, I hope, but don't set people up to hate on it.
Besides, what would they sue over? The FreeBSD team using code that Sun deliberately and explicitly licensed for such things?
It's not Sun you need to worry about, it's NetApp.
Re: (Score:2)
Why? NetApp has been having it's ass handed to it over the patent claims.
Re: (Score:2)
Legal intimidation is a common tactic.
Re: (Score:2)
What's intimidating? In their case against Sun they've had a patent struck down and summary judgements against them saying that Sun doesn't violate others.
Re: (Score:3, Insightful)
What's intimidating?
Being a hobbyist OSS developer and getting hit with a patent infringement lawsuit from a large corporation.
Re: (Score:2)
Re: (Score:3, Informative)
It is not production ready, I know I tested it. The next version should fix those gripes. Patents are what Oracle will sue over.
Re: (Score:2)
How do you think it is not a substantive limitation?
ZFS now only runs on servers. Yeah, they wanted to see it on digital cameras, but in the current market that's not real. Nobody who's currently using ZFS would blanch at the x86_64 requirement.
Sure, there are opportunities that will open up when it's ported further, but they're doing the right thing by getting it out where it'll get the most use.
Re: (Score:2)
Because their are no servers running linux using non-x86_64 CPUs? You had better tell IBM that.
Re: (Score:2)
Because their are no servers running linux using non-x86_64 CPUs? You had better tell IBM that.
If it's important enough for IBM, let them supply the patches for POWER chips. I'd think the x86_64 requirement is mostly for 64 bit addressing, which as far as I know th POWER chips do as well so it shouldn't be that hard given enough interest. Personally I'll wait and see how well this GPL/CDDL module business will work out...
Re: (Score:3, Insightful)
Don't be intentionally dense. The majority of the market that can help these guys refine their code is fine with x86_64.
Re: (Score:3, Interesting)
It's worth mentioning that the latest version of Windows Server (2008 R2) is 64-bit only as well.
And ZFS has always had 64-bit as minimum system requirements for production systems, even on Solaris.
That is, 32-bit is considered okay for limited testing, unsuitable for production use, particularly for use with zpools larger than a few hundred GB in size or so.
If you have a 1TB or larger storage pool with ZFS, you need 2gb of RAM and a 64-bit CPU to have something acceptable and stable. This is tr
Re: (Score:3, Informative)
Because ZFS is not production quality on a 32-bit CPU or with less than at least additional 2GB of RAM available for ARC, even on Solaris where ZFS is most mature. Bare minimum for ZFS: 1Gb RAM, 64bit proc.
If you have a 32-bit CPU or less than 2GB system RAM, use UFS or Ext3, forget about ZFS for such hardware configurations, unless you want to experience pain (system hangs, memory starvation, crashes / Panics due to 32-bit address space squeeze causing fragmentation and ultimately inabi
Re: (Score:2)
FreeBSD already has a fairly good implementation of ZFS. It's other file systems (ext4, xfs, jfs, jffs2...) that it lacks. If FreeBSD ever ditches archaic UFS, expect heads to explode :-)
Re:Good Article (Score:5, Informative)
My phone runs linux and is not x86 of any shape or register size, nor is my workstation, nor are many other machines I have running linux
I can't speak for the Linux version, but ZFS on FreeBSD needs x86-64 for three reasons:
First, and most simply, this is the platform that all of the ZFS developers use, so it is the one that is most tested. This doesn't mean that it won't work elsewhere, it just means that it is not well tested anywhere else.
The second is a performance consideration. ZFS uses a lot of 64-bit arithmetic for computing checksums and so on. On most 32-bit platforms, doing 64-bit arithmetic means that you need to split the operands between two registers, effectively halving the number of GPRs that you have to work with. On x86-32, this basically limits you to 2 registers, which cripples performance - every operation involves some stack spills. This is an x86-specific limitation. On ARM, for example, you have 16 32-bit registers, which can be viewed as 8 64-bit registers for certain instructions. Doing a lot of 64-bit arithmetic on an ARM chip still doesn't generate as much register pressure as even doing 32-bit operations on x86.
The final limitation is memory. ZFS likes to have 600MB or so of kernel memory. On x86, the divide between kernel and userspace memory is typically done using segmentation. The kernel has one segment, marked in the GDT as requiring ring-0 permission to access. When you switch to kernel space, the segment register points to this entry. In userspace, you use other segments (sometimes just one per process, sometimes one for stack, one for heap, and so on, sometimes one for all processes with some churn between them). With other implementations, this is done at the page level, although that's more expensive. The kernel's memory, however, is always mapped into the userspace process's address space - it just isn't always accessible.
The reason for this is that x86 lacks sensible TLB controls. If the kernel's address space were not mapped in this way, then every system call would require a TLB flush, which would impact performance. The more address space that you allocate to the kernel, the less you give to userspace apps. If the kernel has 2GB of address space, userland apps can only have 2GB each. On ARM, each TLB entry is tagged with an ASID. The kernel and userspace programs' address spaces are entirely separate, but transitions between the two don't require a TLB flush because the userspace process can't see entries tagged with the kernel's ASID.
Rather than saying that ZFS requires 64-bit, or requires x86-64, it's more accurate to say that it won't work (well) on x86-32 due to inherent limitations of the platform. That doesn't mean that it won't work well on other 32-bit or 64-bit architectures which are less braindead.
Re: (Score:2)
If FreeBSD 8.1 RELEASE, STABLE and CURRENT are any indication, FreeBSD 9 has a long, long way to go to be the 'home' for much of anything outside of the odd server here and there the base for other projects (pfSense/freeNAS). It certainly is not a 'general purpose' OS - if for no other reason than its woefully lacking (and often unstable) hardware support.
And by 'long way to go' I mean it needs to regress - in stability and commitment to making all the subsystems actually work out the door. The current stat
Hey if Phoronix says it, it has to be true! (Score:4, Funny)
Re: (Score:2)
...or the yahoo toolbar.
(why is oracle pimping the yahoo toolbar, btw, on their java installs? boggle!)
Re: (Score:3, Informative)
They get paid to include that. Just like Microsoft is now paying Verizon to add bing search to phones, and NASCAR pays Sprint to include their apps.
Re: (Score:3, Interesting)
"Port Windows-based games to the Linux platform".
Just sayin'.
Re: (Score:2)
Just sayin'.
ZFS recap (Score:2, Funny)
Re: (Score:3, Funny)
But we're all slashdotters here, so who can read the article to do the recap? I believe this is a belling-the-cat problem.
Re: (Score:2)
Perhaps you'd like a colonel to read it for you?
Re: (Score:3, Informative)
We've heard much about ZFS, but being a slashdotter, I can't recklessly go on and RTFA. So, maybe someone here can recap its main benefits. Maybe a power point slide?
Here's a good PDF on it:
http://hub.opensolaris.org/bin/download/Community+Group+zfs/docs/zfslast.pdf
Here's the PDF being presented by the co-creators, Jeff Bonwick and Bill Moore:
http://blogs.sun.com/video/entry/zfs_the_last_word_in
Three parts, one hour each. Streamable blip.tv as well as a downloadable M4V file.
Two, ten minute videos:
http://www.youtube.com/watch?v=gthel59G56c
http://www.youtub
Can I remove a disk from it yet? (Score:5, Interesting)
http://www.opensolaris.org/jive/thread.jspa?threadID=131604 [opensolaris.org]
http://www.opensolaris.org/jive/thread.jspa?messageID=270957 [opensolaris.org]
Long story short: disk pools in ZFS can only grow, so don't make any mistakes unless you can afford to do a full dump and restore. Sun had been "working on" this for years. Anyone heard any news lately?
Re:Can I remove a disk from it yet? (Score:4, Informative)
The ZFS design makes this very difficult. Btrfs, on the other hand, has supported this feature for a long time, thanks to a nice design feature called backrefs.
Re:Can I remove a disk from it yet? (Score:4, Funny)
not holding my breath, but... (Score:2)
I'm not holding my breath, but god damn...
If this ever happens, my wildest dreams will have come true. Forget the threesome with now-hideous-and-leathery-old-but-once-hot porn stars from my youth! Not having to deal with any more Solaris or FreeBSD for a 'modern filesystems' would be incredible.
Like I said, not holding my breath (or even breathing heavily!). If it happens it'll happen only at the fringes, and poorly.
Re:Freedom ain't free (Score:5, Insightful)
Sun used the CDDL just to make sure Linux never got ZFS. Even that move is not going to save solaris, only open sourcing it earlier would have done that. I say this as a linux user who likes solaris and thinks it will be a shame to see it die. Well I like it once the GNUtools are installed, the solaris versions sucked.
They are both quite open, how free they are some might argue about.
Re: (Score:2)
No, Sun used the CDDL because they hate the restrictions on GPL. The sharing issues go both ways, Sun wanted to keep some ownership. It's not like the BSD license exists just to spite GPL.
Re:Freedom ain't free (Score:4, Insightful)
No, they are a company that exists to make money. Saving Solaris would make them more money. Very simple. Corporations do not hate like that, they only do what they must to maximize profit.
BSD is a fine license, it was created for a real purpose, not to just protect a doomed product.
Re: (Score:2)
Re: (Score:3, Insightful)
It's not hard to be incompatible with the GPL. The GPL prohibits a lot of actions. Good for some, not for others.
That's not the GPL's fault (Score:4, Insightful)
That's not the GPL's fault. It's the fault of the IP lawyers who are dicing permissions exceedingly fine. The GPL is designed to guarantee certain freedoms at the cost of others. It does its job very well, and is well architected with a lot of forethought considering we're only on version three after 21 years. At least one of those two revisions can be blamed not on the faults of the license but on the changing legal and IP environment.
Believe it or not once upon a time if you wrote some code somebody found interesting you just sent it to them. No patents. No copyrights. No approvals from management or legal. You just sent it, happy that someone else might benefit from not redoing the work you'd done once already. The idea of profiting from the derivatives they might make, or the derivatives of the derivatives, was simply not an idea that would occur to a normal person. If you had suggested such a thing at that time we'd have thought it hilarious.
And now I have to point to the onion on my belt, which was the fashion in my day.
Re:That's not the GPL's fault (Score:4, Insightful)
Yes, it is the GPL's fault. The CDDL is a per-file license. It places absolutely no restrictions on what other code can be combined with it in other files.
As the CDDL is deliberately GPL incompatible, had there not been any other issues, one can assume that Sun would have added 'may not be distributed together with GPL licensed code'. The CDDL/GPL incompatibility was on purpose, it was a feature asked for by Solaris engineers. Had the Linux kernel been BSD licensed, the CDDL would have been made incompatible with the BSD license.
Generally, fault implies some form of control over the issue. Under the circumstances, the only party with any control in this case would have been Sun, and as they would have redesigned the license until it was not compatible, it's quite obvious where any 'fault' should be assigned.
And unless the Oracle buyout has changed some attitudes within Sun for the better (heh), it's also quite naive of KQ Infotech to believe that Sun/Oracle would not go after them for violating the point of the license, as opposed to the actual text of the license (assuming any wider distribution). Standing is hardly a necessary prerequisite for a company of Oracles size to grind a small company into dust in the courts (and both Oracle and Sun would have standing as kernel contributors to sue any distributor of ZFS+Linux kernel combo).
Personally I can't say I consider it either a big loss or much to complain about. ZFS was a huge (HUGE) deal for Solaris, considering the painfully anemic storage stack it had in disksuite+ufs, but for any OS with a more modern volume management and file system stack it merely boiled down to a few nice features and some drawbacks, depending on your underlying storage architecture (SAN capabilities, etc).
Re: (Score:2)
Re:Freedom ain't free (Score:5, Informative)
No, Sun used the CDDL because they hate the restrictions on GPL. The sharing issues go both ways, Sun wanted to keep some ownership. It's not like the BSD license exists just to spite GPL.
This is the third time I've seen someone post something to this effect in the past week. I smell a smear campaign. Nonetheless, I'm calling BS here. Daneese Cooper, one of the individuals who helped draft the CDDL, stated that they based the CDDL on the MPL "partially because it is GPL incompatible. That was part of the design when they released OpenSolaris." It was made deliberately GPL-incompatible, but this has nothing to do with 'restrictions' in the GPL.
Re: (Score:3, Informative)
This is the third time I've seen someone post something to this effect in the past week. I smell a smear campaign.
Nonetheless, I'm calling BS here. Daneese Cooper, one of the individuals who helped draft the CDDL, stated that they based the CDDL on the MPL "partially because it is GPL incompatible. That was part of the design when they released OpenSolaris." It was made deliberately GPL-incompatible, but this has nothing to do with 'restrictions' in the GPL.
And Cooper's assertion was reject by Simon Phipps, Sun's Chief Open Source Officer for quite a while (before leaving Oracle in the last few weeks):
http://www.opensolaris.org/jive/message.jspa?messageID=55013#55008
http://en.wikipedia.org/wiki/Common_Development_and_Distribution_License#GPL_incompatibility_controversy
Re: (Score:3, Insightful)
Re:Freedom ain't free (Score:4, Informative)
when it comes to license compatibility issues in general, it is the GPL which is decidedly incompatible with every other license.
That's FUD if I've ever seen FUD. Check out the FSF's list of free software licenses [gnu.org]; there's many licenses that ARE GPL-compatible. Excluding the GNU licenses themselves, there's at least Apache 2.0, Artistic 2.0, Berkeley DB, Boost, Modified BSD, CeCILL, Clear BSD, Cryptix, eCos 2.0, Educational Community 2.0, Eiffel Forum 2, EU Datagrid, Expat, FreeBSD (!), FreeType, iMatix, Independent JPEG Group, imlib2, Intel Open Source, ISC, NCSA, Netscape Javascript, OpenLDAP, Perl 5, PD, Python 2, Python up to 1.6, Ruby, SGI B 2.0, SML/NJ, Unicode, VIM 6.1+, w3c, webm, WFTPL 2, X11, XFree86 1.1, zlib and Zope 2.
And keep in mind that these are *licenses*; in reality, most projects won't even bother making up their own licenses. "Decidedly incompatible with every other license". Sheesh!
some GPL advocates tend to view those who choose a non-GPL license as trying to thwart GNU and/or Linux so they don't have to admit that maybe other licenses have terms and conditions that have their own merit.
Who are those mysterious "GPL advocates" you mention, then? Also, what does this have to do with a situation where Sun really WAS trying to "thwart GNU and/or Linux", by its own admission?
Look, the CDDL isn't a bad license per se, and the FSF page linked above lists it as a free software license, too, if a GPL-incompatible one (it does urge you not to use it for that reason, but hey, this *is* the FSF). But the original point was that Sun wanted to make sure that ZFS etc. would not be available on Linux, and they chose/engineered a GPL-incompatible license specifically to ensure that. You're not even contesting that anymore, so why are you still arguing about the whole thing?
It's a fact. Sun didn't want Linux to get ZFS. Get over it.
Re: (Score:3)
It doesn't sound like there was an agenda against the GPL itself, just the intention to release the code under a license that prevented bundling with Linux.
Re: (Score:3, Interesting)
There's a despair poster, I believe, with a caption along the lines "it could be, your main purpose in life, is to provide a warning to others". (Damn it, the internet made me check [yimg.com] ... "It could be that the purpose of your life is only to serve as a warning to others."
ZFS's purpose was not to be a next generation file system, but to encourage next generation file systems to be built. Free Software has a tendency to get stuck at "good enough" sometimes. And someone has to come along and show that there
Re: (Score:2, Insightful)
Free Software has a tendency to get stuck at "good enough" sometimes.
Not just free software. IE6 has the poster child of getting stuck at "good enough" until it was kicked out of place.
Re: (Score:2, Flamebait)
> ZFS's purpose was not to be a next generation file system, but to encourage next generation file systems to be built. Free Software has a tendency to get stuck at "good enough"
This is a tad ironic considering that a 3rd party had to bail out Sun in this regard.
Solaris was the perfect example of "rut-ware".
Re: (Score:3, Insightful)
I'll be the first to say that ZFS has some shortcomings and limitations.
However, it's like Active Directory is to the workstation/server model of enterprise networking: it does everything. There is nothing else which comes close (or shell we say, 9/10ths of the way) to it in terms of it's "completeness" and feature set. Yes, it has some severe limitations ("Windows only") and shortcomings ("OMG it's a pain to troubleshoot"), with a fairly deep learning curve as well as a limited domain of applicability beyo
Re:Freedom ain't free (Score:4, Insightful)
Instead, btrfs, hammer, etc were developed -- much better, much cleaner file systems.
How can filesystems that don't exist in stable release form yet be "better" than ZFS?
ZFS is far ahead of btrfs, both in terms of stability, features, and usability. Btrfs doesn't have parity RAID, dedupe, or replication yet. These are critical features for large-scale systems. In short, it isn't even close to ZFS. ZFS is also "cleaner" in my opinion, in both design and UI. Oracle funding most btrfs development also raises a question of btrfs momentum now that they own ZFS and Solaris.
Re: (Score:3, Informative)
Comparing hammer to ZFS is also a bit silly. Hammer was developed precisely because ZFS did not solve the problem that DragonflyBSD wanted solved. ZFS is designed for large SANs controlled from a central server. Hammer is designed to allow you to treat every disk on a network as part of the same storage pool. They are diametrically opposed objectives, and a filesystem designed to do both would need to either make painful compromises or have so much variation in code paths that it would effectively be tw
Re: (Score:3, Funny)
What could be better than taking a sledgehammer to a disk drive without causing problems?
Shooting it with a .45?
Open Source != Free Software (Score:2, Informative)
Re:Open Source != Free Software (Score:4, Insightful)
I don't know if that's true. I know you probably can't redistribute the kernel with the CDDL bits but you can redistribute them separately (CDDL = Common Development and Distribution License). Then all you have to do is make sure that your software (or customer) installs the right bits and then you can get a pretty decent NAS box.
Besides the legal issues, I would love to see them tackle the technical issues. ZFS itself is very clean in code, very well documented and pretty simple once you get down to the wire. The issue (and selling point) is going to be performance and upkeep and for commercial implementations support. If the upkeep is going to be similar to BSD's implementation (several versions behind) or the performance as bad as FUSE, people are just going to stick to OpenSolaris (or one of it's commercially supported decendants like Nexenta).
Re: (Score:2)
Re: (Score:3, Interesting)
I know you probably can't redistribute the kernel with the CDDL bits but you can redistribute them separately
Not necessarily. I know programmers like to interpret the law as if it were a computer program--rigidly, and with no room for interpretation--but it doesn't necessarily work that way. The law will consider your intent, and if your intent is to deliver a derivative work, then forcing the customer to obtain the pieces and glue them together won't necessarily let you off the hook.
This came up most famously with Objective C. Originally, Steve Job's NeXT made a proprietary front-end to GCC, and they made their
Re: (Score:2)
Re:Open Source != Free Software (Score:4, Insightful)
This is both Open and Free, just not quite as free as Stallman would like.
CDDL licensed code can be freely distributed and modified, so long as it is compiled with a compatible license.
This is why BSD has no issues with including ZFS. The BSD license is less restrictive than the GPL.
Re: (Score:2)
Where has Stallman ever said that it's not as free as he would like? I know the FSF disapproves of licenses that aren't GPL-compatible, but not on the grounds that they're not free enough. They simply don't like code that they can't use for their own projects--and who can blame them? I don't either, and I use a wide variety of licenses, including ones that the FSF isn't necessarily thrilled with.
Re: (Score:2)
This is both Open and Free, just not quite as free as Stallman would like.
No, the problem is that both licenses have the same kind of copyleft clause:
CDDL: "You may not offer or impose any terms on any Covered Software in Source Code form that alters or restricts the applicable version of this License or the recipients rights hereunder."
GPLv2: "You may not impose any further restrictions on the recipients' exercise of the rights granted herein."
Anything that is a requirement in one would be seen as a restriction in the other, the only way it would work was if one license was a st
Re: (Score:3, Informative)
They don't prevent use. They prevent redistribution as part of the whole.
I can download, build, and install fglrx (which is completely non-free) or this ZFS module. I just can't distribute either module linked into the kernel.
Re: (Score:2)
Re: (Score:3, Insightful)
The GPL prevents redistribution of more free combinations as well.
...more free combinations which allow further redistribution of less free combinations, to be more exact.
Without this, GPL would be rather pointless. If somebody wants to keep their code free, and by extension, allow all future users of the code have certain freedoms that come with having the source code (what freedoms exactly, depends on GPL version, due to tivoization [wikipedia.org]), then that's what is needed.
Freedom (of any kind) is not black and white thing, nor is it one-dimensional scale.
Freedom to take freedom aw
Re: (Score:2)
BTRFS will end ZFS if Oracle does not kill that too.
If it comes out and works well (Score:5, Informative)
Seems a little early to be putting faith in that. It's feature list looks good, on par with other modern desktop file systems like HFS+ and NTFS. However it is currently unstable. When will that be fixed? Who knows? Maybe it moved full steam ahead and we have a stable, capable file system next month. Maybe the project loses steam and languishes and 4 years from now it is still "unstable" and "coming soon."
You can't really say how well it'll work until there is stable code to test. Remember designing a file system isn't the real hard part. I'm not saying it is trivial work or that it is unimportant but it is by far the easier part of all this. You can write out a specification that sounds great on paper, but then you have to implement it. That is the much harder part. You have to make it fast, stable, not corrupt data, able to do everything it should and so on.
This is part of the reason why NTFS on Linux has been so tricky. It is actually pretty well documented in the Windows Internals book, and other places, but it is a complex file system. FAT, on the other hand, is real simple and thus not hard to implement.
As an example you can look at driver sized. The NTFS driver in Windows is 1.6MB. The FAT driver, on the other hand which supports multiple versions of FAT, is only 200k. The NTFS kernel driver is one of the very largest in the system, only the ATi video driver (much larger) and TCP/IP stack (a bit larger) are bigger than it on my system.
So we'll see what happens with btrfs. As of late, there's not been much activity. The last version update was June 2009. Maybe they are rolling up final testing for production release, or maybe things have slowed down and release is not near. We'll just have to wait and see, but it is foolish to believe this will be the Next Big Thing(tm) at this point.
Re:If it comes out and works well (Score:5, Informative)
Re: (Score:3, Insightful)
Well when it comes to filesystems, I wouldn't want to run one on any production machine until the developers say they are stable. Bad FS driver can equal file system corruption and data loss.
The problem with testing early release stuff is that it is not necessarily representative of the final product. It can, of course, be too slow because of lack of optimization. However it can also be too fast. What do I mean by that? I mean perhaps during implementation, they discover that more extensive checks and proce
Re: (Score:3, Informative)
Re: (Score:2, Insightful)
You ready to try it on a live Oracle database server with a 10TB filesystem that averages at least say 5,000 DB transactions per second? :)
Re: (Score:2)
"Seems a little early to be putting faith in that. It's feature list looks good, on par with other modern desktop file systems like HFS+ and NTFS."
LOLWUT? NTFS is about as far from 'modern' file systems as ext2/ext3.
NTFS has a lot of problems with performance (like a tendency to fragment seemingly while the HDD is still en-route in plastic package). It doesn't support deduplication, symbolic links to files (yep, it's done above the VFS layer in Windows Vista), no support for RAIDs, no support for dynamic re
Re: (Score:3, Informative)
"*symbolic links to files, incorrect NTFS has supported reparse points since Windows 2000"
Incorrect. Reparse points apply only to directories, not files.
"*no support for RAIDs, incorrect Server versions support RAID 0, 1, and 5"
On block level. No filesystem support, like in BTRFS or ZFS.
"*no support for dynamic resizing, incorrect Windows 2003 added support for dynamic growth for non-system/non-boot volumes, 2008 added dynamic grow and shrink for all volumes."
Only for 'dynamic' disks which are undocumented
Re:If it comes out and works well (Score:5, Interesting)
"* *Actual* performance problems due to fragmentation - outside of a few corner cases - are basically nonexistant. "
Yep. That's why I have to run defragmenter on our build server every week...
Also, Windows is notoriously slow with file operations. It's not directly related to NTFS, but more to extremely inefficient VFS stack.
"* Can you explain what you mean by "it's done above the VFS layer" ? Surely you're not trying to argue symlinks and shortcuts are the same thing ? "
http://neosmart.net/blog/2006/vista-symlinks-revisited/ [neosmart.net]
"* RAID is handled at the block device level, not the filesystem level (and many, many people believe putting RAID into the "filesystem" is an architecturally bad thing, so that's hardly something it can be plainly criticised for)."
However, filesystem-level RAIDs have a lot more functionality than block-level RAIDs. Look at ZFS or BTRFS.
"* Do you have a source for up-to-date benchmarks ?"
I have my own set of benchmarks. Well, NTFS on Windows is almost always slower (and quite often like 100 _times_ slower) than Linux filesystems.
http://rsdn.ru/File/37054/benchmark.zip [rsdn.ru] - this is the source.
http://rsdn.ru/forum/philosophy/1710544.1.aspx [rsdn.ru] - this is a post with benchmark results (in Russian, sorry - I can translate if you have any questions)
http://rsdn.ru/forum/philosophy/1712431.aspx [rsdn.ru] - this post contains this benchmark, slightly adapted.
I regularly re-run these tests. So far, Windows is only getting slower compared to Linux.
I've recently created a multithreaded version of this test. Well, let's say that NTFS sucks so badly, that it's hard to understand how MS has managed to achieve this.
Re: (Score:3, Insightful)
Not that I doubt you, but have any references to verify that claim please?
Re: (Score:2)
Re: (Score:3, Informative)
Note that Btrfs does not yet have a fsck tool that can fix errors. While Btrfs is stable on a stable machine, it is currently possible to corrupt a filesystem irrecoverably if your machine crashes or loses power on disks that don't handle flush requests correctly. This will be fixed when the fsck tool is ready.
https://btrfs.wiki.kernel.org/index.php/Main_Page [kernel.org]
Re: (Score:2, Insightful)
If you think LVM + ext3 is better than NTFS sorry that speaks only to ignorance and possible Linux zealotry, not to knowledge. I don't care if you like MS, their file system is high class. It offers good performance and a very wide feature set, and they update it all the time. They don't rename it, but NTFS has been improved with each version of Windows.
No shame in not knowing about file systems, it is rather esoteric, but then please don't go shooting off at the mouth about how $your_chosen_platform has th
Re: (Score:3, Funny)
> If you think LVM + ext3 is better than NTFS sorry that speaks only to ignorance and possible Linux zealotry, not to knowledge.
Too bad you spent all of that space and didn't actually come up with any real reasons to justify this position.
Re: (Score:2)
They don't rename it, but NTFS has been improved with each version of Windows ... No shame in not knowing about file systems, it is rather esoteric, but then please don't go shooting off at the mouth about how $your_chosen_platform has the bestest FS and everything else sucks.
1. Sorry, mate, NTFS does indeed have unique version numbers.
2. Like most Micorosft apologists, you're spending considerable time and effort going on about "features", an approach that's more appropriate to a marketing brochure than a
Re: (Score:2)
Re: (Score:2)
When did it start letting people do the equivalent of "rm /bin/rm"?
Re: (Score:3, Informative)
NTFS doesn't do COW, but it's had snapshotting for a while under the name "volume shadow copy". This was added in XP or 2003, and even given somewhat of a UI in the form of "previous versions" in Vista.
Re: (Score:3, Insightful)
So you are suggesting I can freeze IO to the machine, then run a snapshot command on NTFS?
I would be glad to hear it.
Re: (Score:3, Informative)
The whole point of snapshots is that you don't freeze the IO. The snapshot service provides you a, well, snapshot of how things were at the moment it was requested, and maintains that snapshot even as other applications keep writing data. It's roughly similar to MVCC, only the units are FS blocks, not database records.
Re:If it comes out and works well (Score:5, Informative)
So you are suggesting I can freeze IO to the machine, then run a snapshot command on NTFS?
I would be glad to hear it.
The Volume Shadow Service (VSS) is always running (by default). Backup utilities - including the ones which come with Windows - use VSS to create a snapshot and perform backup from that point in time. It doesn't freeze IO; rather it goes to copy-on-write.
On server versions you can also create snapshots interactively by using the vssadmin tool.
Shares can be set up to create a shadow copies multiple times per day. This is not copy on every write - but it *is* copy on write once a block is part of a snapshot. Any client (plugin needed for XP, IIRC) can display previous versions which are available snapshots.
VSS actually goes beyond NTFS integration (which is probably why it is a service and not just a NTFS feature). Certain applications - e.g. Exchange, SQL Server and Hyper-V - also integrate with VSS. Instead of VSS operating directly on e.g. SQL Server files, it integrates with the server to create a snapshot for the database files. During restore the system knows how some applications took part in the shadow copy. This ensures that I can correctly restore *all* the files needed to bring a SQL server database back to a certain point-in-time. It also allows the SQL server to prune the log automatically.
I have a Server2008R2 which has several Hyper-V images (development and testing). When I perform a backup of the server, VSS interacts with Hyper-V to perform backup of the virtual machines as well. A Server2003 which hasn't been set up to support VSS is actually "hibernated" by Hyper-V/VSS - then backed up - then brought back into running state. That could be considered "freezing IO", I suppose.
Re: (Score:2)
Re: (Score:2)
In Windows 2003 (Score:3, Informative)
It was deployed to desktops, and on by default, in Windows Vista/7. It does copy on write and maintains old snapshots of files automatically. On the server side, there is some more management of this if you like. This snapshotting feature is also used by backup utilities to do hot backups. Ghost and TrueImage can image a running system using it. They can snapshot the state for backup and new data can be committed while they work, without messing with anything. Works great. That is also independent of the ma
Re: (Score:2)
Ghost and TrueImage can image a running system using it.
Having worked on True Image, I think I need to qualify that. I don't know about the recent versions (though I doubt it would have changed), but back in the day it was only using VSS API to notify applications that a snapshot for backup purposes is taking place, so that they can bring themselves to consistent state before it happens (this is most useful for database servers). However, actual snapshotting code in True Image is entirely its own in-house developed kernel driver.
Re: (Score:2)
A long time ago.
http://axelilly.wordpress.com/2009/01/15/creating-vss-snapshots-on-windows-2003-server/ [wordpress.com]
http://en.wikipedia.org/wiki/Shadow_Copy [wikipedia.org]
FreeBSD had it around '05, Windows 2003 introduced COW snapshots. The new Win 7 and Server 2008 offer diskshadow, a decent cli interface to manage them.
EvanED is apparently not familar with VSS. I don't know of UI to VSS which is not 3rd party one.
If you've used any of the COW snapshot capable FS's, you know why LVM snapshots are inadequate.
Re: (Score:2)
Re: (Score:2)
Re: (Score:3, Insightful)
ZFS has becoming vapor ware since apple announced snow kitty wasnt gunna support it.
I do not think that word means what you think it means.
Re: (Score:2)
Mount a nice fast SSD as swap would be a good start. Then point your /tmp to a tmpfs.
Re: (Score:2)
Mount a nice fast SSD as swap would be a good start. Then point your /tmp to a tmpfs.
That's not even remotely close to the same thing.
Re: (Score:2)
I would honestly then say mount it as /. Seems the easiest way to do something similar to what you want. Not exactly the same, but with 32GB SSDs being very affordable not a huge hurdle either.
Re: (Score:3, Funny)
I seriously doubt that mounting /. As your filesyatem will help in any way. You'd end up just reading and commenting on each post and forget TFA altogether.
Re: (Score:2)
Re: (Score:2)
Re: (Score:3, Interesting)