Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Native ZFS Is Coming To Linux Next Month

timothy posted more than 4 years ago | from the is-it-as-magical-as-advertised? dept.

Oracle 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.'"

Sorry! There are no comments related to the filter you selected.

Freedom ain't free (0, Flamebait)

binarylarry (1338699) | more than 4 years ago | (#33399662)

How open are the two open source licenses if they prevent you from using the software?

Re:Freedom ain't free (0)

Anonymous Coward | more than 4 years ago | (#33399700)

They don't prevent you from using it. However, the GPL prevents redistribution of less free combinations. You can use anything GPL internally (in your home or inside your organization) combined with anything non-free which you like, you just are not licensed to redistribute it to external organizations or people.

Re:Freedom ain't free (4, Insightful)

h4rr4r (612664) | more than 4 years ago | (#33399706)

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:Freedom ain't free (1)

outZider (165286) | more than 4 years ago | (#33399794)

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 (3, Insightful)

h4rr4r (612664) | more than 4 years ago | (#33399820)

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:Freedom ain't free (1)

symbolset (646467) | more than 4 years ago | (#33400034)

No, Sun used CDDL to be deliberately incompatible with the GPL because they knew the good bits would get assimilated into Linux if they didn't.

Re:Freedom ain't free (0)

Anonymous Coward | more than 4 years ago | (#33400052)

they knew the good bits would get assimilated into Linux

Linux is Borg now?

Re:Freedom ain't free (5, Informative)

coerciblegerm (1829798) | more than 4 years ago | (#33400132)

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:Freedom ain't free (3, Interesting)

larry bagina (561269) | more than 4 years ago | (#33399914)

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 is a better way. Competition is good. Sometimes it's internal (gcc vs egcs), sometimes it commercial (CVS vs perforce and bitkeeper).

What if ZFS was GPL? What if it went into Linux? It might get incremental tweaks, but it would stagnate at "good enough". Instead, btrfs, hammer, etc were developed -- much better, much cleaner file systems.

ZFS has some cute tricks. What could be better than taking a sledgehammer to a disk drive without causing problems? But ultimately, ZFS would hold linux back.

Re:Freedom ain't free (2, Insightful)

EvanED (569694) | more than 4 years ago | (#33400050)

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:Freedom ain't free (1, Flamebait)

jedidiah (1196) | more than 4 years ago | (#33400198)

> 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:Freedom ain't free (1)

fucket (1256188) | more than 4 years ago | (#33400322)

Thanks for linking to the thumbnail. I'll take your word for it that that's what it says.

Open Source != Free Software (2, Informative)

cpicon92 (1157705) | more than 4 years ago | (#33399712)

It's open source in the sense that the source is open. Free to view, and free to use as long as you don't distribute it.

Re:Open Source != Free Software (3, Insightful)

guruevi (827432) | more than 4 years ago | (#33399776)

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:Open Source != Free Software (1, Informative)

mysidia (191772) | more than 4 years ago | (#33399888)

I don't know if that's true. I know you probably can't redistribute the kernel with the CDDL bits

Just like you can't distribute a Linux distribution and include nVidia drivers? Tell that to the distros [wikipedia.org] .

But I think it may fall on death ears. Or you might hear back a reminder that mere aggregation on a storage medium is exempted in the GPL... as long as the non-free package is separated from the GPL'd package.

Re:Open Source != Free Software (1)

straponego (521991) | more than 4 years ago | (#33400136)

Death ears? Be very, very quiet... the death ears will get you.

Re:Open Source != Free Software (1)

mysidia (191772) | more than 4 years ago | (#33400154)

I meant deaf ears, mainly because the distros. ignored whatever complaints they got and continued to include non-GPL-compatible drivers.

Re:Open Source != Free Software (1)

hedwards (940851) | more than 4 years ago | (#33400270)

The main reason why the BSD implementation is several versions behind is that ZFS has been under rapid development lately. Last time I checked, FreeBSD had the important features ported, with a few nice to haves not yet available. The important work was on the performance end of things and squashing OS specific bugs. This won't remain like this forever, at some point ZFS will mostly be fleshed out and in need of only a small number of tweaks, at which point I doubt other OSes will be that far behind. A filesystem isn't terribly useful if only one specific release can properly utilize it.

Re:Open Source != Free Software (2, Interesting)

Xtifr (1323) | more than 4 years ago | (#33400284)

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 customers get the rest of GCC for themselves. When the FSF objected, Steve consulted his lawyers, and decided to release the Objective-C front end source, which is why GCC includes Objective-C today.

Of course, the laws may be different in India, but I wouldn't touch this module without consulting with my lawyers first, since I'm in the US.

Re:Open Source != Free Software (0)

Anonymous Coward | more than 4 years ago | (#33399832)

CDDL is a free software license, so what's your point?

Reference: http://www.gnu.org/licenses/license-list.html#GPLIncompatibleLicenses

Re:Open Source != Free Software (3, Insightful)

quercus.aeternam (1174283) | more than 4 years ago | (#33399922)

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:Open Source != Free Software (1)

Xtifr (1323) | more than 4 years ago | (#33400296)

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:Freedom ain't free (2, Informative)

stinerman (812158) | more than 4 years ago | (#33399720)

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:Freedom ain't free (1)

hedwards (940851) | more than 4 years ago | (#33400286)

Precisely, FreeBSD for a long time (Possibly still) shipped with two math emulators, one which was compiled in by default was kind of bogus but BSD licensed, the other which only included source was GPLed, but happened to work better. It was a bit of a pain, but ultimately worked, people didn't usually redistribute custom kernels anyways, so the requirements of the GPL weren't particularly restrictive on the typical user and everybody could have what they wanted.

Good Article (5, Insightful)

bill_mcgonigle (4333) | more than 4 years ago | (#33399722)

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 (4, Insightful)

h4rr4r (612664) | more than 4 years ago | (#33399744)

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:Good Article (1)

Monkius (3888) | more than 4 years ago | (#33399788)

Sadly, yes, there is every reason to expect it.

Re:Good Article (1)

Just Some Guy (3352) | more than 4 years ago | (#33399858)

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:Good Article (2, Informative)

bill_mcgonigle (4333) | more than 4 years ago | (#33399904)

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:Good Article (1)

Lunix Nutcase (1092239) | more than 4 years ago | (#33400064)

Why? NetApp has been having it's ass handed to it over the patent claims.

Re:Good Article (1)

bill_mcgonigle (4333) | more than 4 years ago | (#33400150)

Legal intimidation is a common tactic.

Re:Good Article (1)

Lunix Nutcase (1092239) | more than 4 years ago | (#33400176)

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:Good Article (2, Insightful)

bill_mcgonigle (4333) | more than 4 years ago | (#33400200)

What's intimidating?

Being a hobbyist OSS developer and getting hit with a patent infringement lawsuit from a large corporation.

Re:Good Article (1)

hedwards (940851) | more than 4 years ago | (#33400220)

It is, however, it's not BSD code, it's a port of the code from Sun. Well, now belonging to Oracle, I'm not sure how they could sue the FreeBSD people without first filing suit against Oracle for infringing on whatever patents they claim are relevant.

Re:Good Article (1)

mysidia (191772) | more than 4 years ago | (#33400180)

That doesn't stop them from pursuing more claims against targets they perceive to be vulnerable. If they think they can get a quick win or quick settlement with BSD folks (to stop distributing), they will do so, and then use it as leverage to attempt to reach a settlement with Oracle.

Probably requiring Oracle to get out of the storage business and stop making ZFS available to ISVs/partners to build NASes with, or as open source

Re:Good Article (2, Informative)

h4rr4r (612664) | more than 4 years ago | (#33399944)

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:Good Article (1)

mysidia (191772) | more than 4 years ago | (#33400228)

ZFS is just about as "production ready" on the latest stable release of FreeBSD, as BtrFS is "Production Ready" on the latest Linux kernel release.

Re:Good Article (1)

bill_mcgonigle (4333) | more than 4 years ago | (#33399880)

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:Good Article (1)

h4rr4r (612664) | more than 4 years ago | (#33399948)

Because their are no servers running linux using non-x86_64 CPUs? You had better tell IBM that.

Re:Good Article (1)

Kjella (173770) | more than 4 years ago | (#33400042)

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:Good Article (2, Insightful)

bill_mcgonigle (4333) | more than 4 years ago | (#33400138)

Don't be intentionally dense. The majority of the market that can help these guys refine their code is fine with x86_64.

Re:Good Article (3, Interesting)

mysidia (191772) | more than 4 years ago | (#33400258)

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 true whether you used Solaris or BSD.

I consider it a good thing that the person porting to Linux is actually enforcing the basic 64-bit requirement. Maybe fewer people who don't read docs and 'system requirements' sheets will get burned that way, by not noticing that "32-bit is not suitable for enterprise use", and say ZFS on Linux 'sucks', because they screwed up basic configuration and deployment requirements ?

Re:Good Article (0)

Anonymous Coward | more than 4 years ago | (#33400082)

If FreeBSD ever gets a good ZFS implementation expect lawsuits to commence.

This is a ZFS system
# uptime
  9:38PM up 750 days, 5:19, 1 user, load averages: 0.16, 0.09, 0.08
# uname -sr
FreeBSD 7.0-STABLE

Re:Good Article (3, Informative)

mysidia (191772) | more than 4 years ago | (#33400212)

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 inability to allocate ARC efficiently).

Re:Good Article (1)

arth1 (260657) | more than 4 years ago | (#33400292)

If FreeBSD ever gets a good ZFS implementation expect lawsuits to commence.

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 (0)

Anonymous Coward | more than 4 years ago | (#33400244)

I disagree with some of his subjective claims

That's funny, I disagree with some of his objective claims. For example, the claim that native ZFS is coming to linux next month...

who cares?! (0, Flamebait)

Dr.D.IS.GREAT (1249946) | more than 4 years ago | (#33399746)

ZFS has becoming vapor ware since apple announced snow kitty wasnt gunna support it.

I guess we could use reiserFS...

just a thought

Re:who cares?! (1)

h4rr4r (612664) | more than 4 years ago | (#33399754)

BTRFS will end ZFS if Oracle does not kill that too.

If it comes out and works well (5, Informative)

Sycraft-fu (314770) | more than 4 years ago | (#33399846)

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 (-1, Offtopic)

h4rr4r (612664) | more than 4 years ago | (#33399878)

BTRFS blows NTFS out of the water. LVM + ext3 does more than NTFS. HFS+ I do not know much about.

Re:If it comes out and works well (2, Insightful)

UnknownSoldier (67820) | more than 4 years ago | (#33399964)

Not that I doubt you, but have any references to verify that claim please?

Re:If it comes out and works well (1)

tepples (727027) | more than 4 years ago | (#33400240)

Another Slashdot user recommends against using "citation needed" in a Slashdot post [slashdot.org] . Apparently, Slashdot users are supposed to first list what they tried in Google. Something like "I tried ext3 vs ntfs in Google but couldn't tell what was reliable" might help.

Re:If it comes out and works well (2, Insightful)

Sycraft-fu (314770) | more than 4 years ago | (#33399966)

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 the bestest FS and everything else sucks.

Linux actually is a good deal behind in the FS game, which is one of the reasons there's so much interest in btrfs or ZFS. Linux could use a more modern file system for many tasks.

Finally Btrfs doesn't "Blow NTFS out of the water," because btrfs cannot be used on production systems. NTFS is used on basically every production Windows system, desktop or server, in the world. Btrfs is still under development. It may be a far superior file system when it comes out, but it isn't out yet. You can't set it up and use it and expect all your data to be ok.

There's a big difference between something under development and something in production. IE9 doesn't "Blow Firefox 3 out of the water," either. In theory it will, I've tried the test builds and its rendering is amazingly fast because of the hardware acceleration. However, it isn't a stable, release browser. It is just a test. How it does in realty when it is released, and equally important how FF is doing at the same time, remain to be seen. Making any claims of it at this point would be extremely premature.

Re:If it comes out and works well (0)

Anonymous Coward | more than 4 years ago | (#33400012)

FAT32 is better than NTFS.

Ever try to use an NTFS-formatted flash drive in anything that's not a PC?

QED

Re:If it comes out and works well (1)

tepples (727027) | more than 4 years ago | (#33400266)

Parent post supports only the assertion "NTFS is better on PCs; FAT32 is better on all other devices." If you'd believe the Windows XP/Vista/7 formatter, FAT32 is aptly named: Microsoft believes that compatibility with non-PC devices isn't worth the performance drop on volumes bigger than 32 GB.

Re:If it comes out and works well (0, Troll)

h4rr4r (612664) | more than 4 years ago | (#33400038)

When did NTFS get snapshotting? A rather simple feature.

Re:If it comes out and works well (1)

h4rr4r (612664) | more than 4 years ago | (#33400058)

When did it start letting people do the equivalent of "rm /bin/rm"?

Re:If it comes out and works well (3, Informative)

EvanED (569694) | more than 4 years ago | (#33400066)

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:If it comes out and works well (2, Insightful)

h4rr4r (612664) | more than 4 years ago | (#33400094)

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:If it comes out and works well (1)

EvanED (569694) | more than 4 years ago | (#33400174)

No need to actually freeze I/O; the volume shadow service will ensure that the snapshot is taken at a consistent time.

I've not actually used this feature, so I spent a bit of time just now looking around for how to actually activate it. It sounds like it's easy if you have a server edition of Windows ("vssadmin create shadow /for=c:"), but I'm not sure how you can trigger a copy manually on non-servers. It's possible that creating a system restore point does exactly what you want, but I'm not sure. You can tell it to make snapshots automatically (for use with the "previous versions" feature), but I don't see a way to specify the interval, just stuff like the maximum amount of space to use.

Re:If it comes out and works well (2, Informative)

shutdown -p now (807394) | more than 4 years ago | (#33400226)

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 (4, Informative)

benjymouse (756774) | more than 4 years ago | (#33400298)

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:If it comes out and works well (0)

Anonymous Coward | more than 4 years ago | (#33400314)

Start
Run
cmd
vssadmin

Re:If it comes out and works well (0)

Anonymous Coward | more than 4 years ago | (#33400340)

http://technet.microsoft.com/en-us/library/cc754968%28WS.10%29.aspx - vssadmin has existed since Windows XP

Re:If it comes out and works well (1)

EvanED (569694) | more than 4 years ago | (#33400100)

NTFS doesn't do COW

I think I misspoke here; volume shadow copy can do COW.

What NTFS doesn't do is the ZFS "always allocate a new block when writing" thing, and that's what I was thinking of when I wrote that.

Re:If it comes out and works well (1)

mysidia (191772) | more than 4 years ago | (#33400348)

I/O is only frozen for 10 seconds. I'm sure Microsoft has an article about this [microsoft.com]

Yes, one method shadow copy uses is COW. Shadow copy is not NTFS.

It is a volume-level mechanism that lies under the NTFS file system. There is an area for files, and there is a volume shadow copy region on each volume.

There are two methods shadow copy can be implemented, one involves re-directing writes to the reserved region. Copy on write is the other option.

In Windows 2003 (2, Informative)

Sycraft-fu (314770) | more than 4 years ago | (#33400190)

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 maintaining of old versions so you can shut that down if you like and still do snapshots for backups.

Re:In Windows 2003 (1)

shutdown -p now (807394) | more than 4 years ago | (#33400248)

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:If it comes out and works well (2, Funny)

jedidiah (1196) | more than 4 years ago | (#33400214)

> 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:If it comes out and works well (1)

value_added (719364) | more than 4 years ago | (#33400262)

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 technical discussion. So while NTFS may have lots of "features", so do most file systems. What matters is the implementation, and how that translates to the real world. On that count, few people are as impressed as you seem to be.

Re:If it comes out and works well (-1, Troll)

Anonymous Coward | more than 4 years ago | (#33400332)

Wow. What a rant from a raving idiot. Last I checked, windows users were still having to 'defrag the drive'. WHAT THE FUCK! That problem was solved in 1967. No one has had to do that since Goldwater was trying to get elected. You claim NTFS is new? WHAT? Are you an idiot? How extremely little do you really know about file systems? Linux has more file systems than ANY OTHER PLATFORM. Linux can support all the latest and greatest NTFS (which is really just a cheap knock off of HPFS which IBM gave to microsoft before microsoft stabbed IBM in the back). If you knew anything at all about Linux or Unix, you would know why Linux can support more kinds of file systems than anyone (including NTFS). There are a lot of filesystems much better than NTFS. It goes beyond B trees, B+ trees, B* trees, dancing trees and the like. It goes beyond whether the basic structure in the tree can be an overloaded object which can take many forms (such as directory, file, socket, pipe or link). Why am I even going on about this. YOU CLEARLY DON"T KNOW ANYTHING about file systems. You've never made one, and all you know is what the juice dispensers at mickeysoft feed you. You are on the juice. You have no brain!

Re:If it comes out and works well (4, Informative)

Christophotron (812632) | more than 4 years ago | (#33399990)

BTRFS is not that unstable really.. I have been running for a few months now, since the on-disk file structure was finalized. it's in a raid 1 configuration across 2 300gig drives on one of my home servers and it hasn't had a hiccup yet, even with lots of file i/o. i think it would like more than the CPU and RAM I gave it, but its still less resource intensive than ZFS. AFAIK ZFS would not even run on that machine due to the 32 bit processor and only 512mb of RAM. Some of the features are not implemented yet but it is certainly stable enough to test..

Re:If it comes out and works well (2, Insightful)

Sycraft-fu (314770) | more than 4 years ago | (#33400152)

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 processing of certain kinds is necessary to maintain data integrity. This ends up slowing things down.

I'm not saying that will happen, just saying it can, hence I don't like tests until the final version is there and everything is in it.

Also the other problem I have is that you don't know if something will ever reach completion. I've seen OSS projects that rocketed to a stable release in record time, I've seen them proceed slow and steady to a release, and I've seen them peter out and stagnate. You really don't know how it is going to be until it comes out.

I remember when at work we were trying to decide on an e-mail client to recommend to users (I work in an academic environment so we can't mandate things). None of us were that happy with Outlook for IMAP (though in 2010 it has gotten a lot better). I was using Thunderbird and found it to be acceptable. Since it was free and worked, I said lets do that. My boss wanted to wait for another one, a project he'd been watching for some time (can't remember the name). Was going to be awesome when it released... Ya well that day never came. They kept talking and talking, but no releasing of final code. Maybe some day it will be awesome, but right now Thunderbird is 100% more awesome since it is out, running and useful.

I have high hopes for btrfs because our large, inexpensive, non-backed up storage (our backed up storage is NetApp) at work is Linux and it could use something better than ext3. However I won't say "Yes this is a good FS," until it is a released product and I've seen how the final thing works.

Re:If it comes out and works well (2, Informative)

Christophotron (812632) | more than 4 years ago | (#33400318)

I agree. BTRFS is definitely not ready for production or for storage of anything important that is not backed up elsewhere. It has known bugs, like for example the reported free space on a raid 1 will show the total disk size and not the actual free space, so it may be dangerous to fill the array too close to 100% (shown as 50% in df). It is unclear when (or if) it will be ready, but it is being worked on -- I've seen updates for the userland tools in Debian testing, and the newer kernels have updates for the fs driver. The bug I mentioned is fixed in 2.6.33, I believe. I was only countering the argument that it is too unstable even to test it out. That is untrue. Heck, even Linus Torvalds reportedly uses BTRFS as the root filesystem on one of his laptops.

Re:If it comes out and works well (1)

Cyberax (705495) | more than 4 years ago | (#33400164)

"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 resizing, etc. It's also SLOW. NTFS is an old filesystem, it was conceived in the beginning of 90-s when journaling was the state of the art. But now the state of the art has moved far far away.

NTFS driver is so large because the internal structures of NTFS are fairly complex. NTFS kernel driver is also not the largest one, XFS is by far larger. ext4 is comparable.

Re:If it comes out and works well (1)

EvanED (569694) | more than 4 years ago | (#33400302)

...symbolic links to files (yep, it's done above the VFS layer in Windows Vista), no support for RAIDs...

To be fair, if these are done in another layer (as they are, if we take your word for it), then you can hardly count it as a strike against the file system.

I'd definitely like to see Windows get a ZFS-like file system, and I do think NTFS is comparable to ext3 etc. (Each has a benefit or two over the others, e.g. ext3 doesn't store a file creation time and Windows has Transactional NTFS, but at least NTFS on Windows has some obnoxious file naming limits and locking problems.) That said, NTFS beat the pants off of the file systems available in 2.4.

Re:who cares?! (3, Insightful)

EvanED (569694) | more than 4 years ago | (#33399784)

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:who cares?! (-1, Troll)

Anonymous Coward | more than 4 years ago | (#33399822)

Snow Kitty? Jealous because you can't afford a real computer?

Re:who cares?! (-1, Troll)

Anonymous Coward | more than 4 years ago | (#33399862)

Jealous because you can't afford a real computer?

So that's what they call buying outdated hardware for almost twice the normal price?

Using SSD as an HD cache / ZFS L2ARC (0)

Anonymous Coward | more than 4 years ago | (#33399758)

I hope this idea http://blogs.sun.com/brendan/entry/test also works with the port (and will be working with btrfs one day!?):

Transparently adding an SSD into the slow-HD-to-fast-CPU-register cache hierarchy. Are there ways to make something like this work with current FSes?

Re:Using SSD as an HD cache / ZFS L2ARC (1)

h4rr4r (612664) | more than 4 years ago | (#33399764)

Mount a nice fast SSD as swap would be a good start. Then point your /tmp to a tmpfs.

Re:Using SSD as an HD cache / ZFS L2ARC (1, Interesting)

Anonymous Coward | more than 4 years ago | (#33399890)

I hope this idea http://blogs.sun.com/brendan/entry/test [sun.com] also works with the port (and will be working with btrfs one day!?):

Transparently adding an SSD into the slow-HD-to-fast-CPU-register cache hierarchy. Are there ways to make something like this work with current FSes?

Mount a nice fast SSD as swap would be a good start. Then point your /tmp to a tmpfs.

My swap is practically always empty, so that would not help much. And how much stuff gets written into /tmp? Mine is 839kiB right now, that is neglible and probably in RAM cache already. What I want cached are all the binaries and libs linked from there. Then starting any app for the second time would be near-instant and this cache would survive reboots and be one or two orders of magnitude larger than RAM.

Re:Using SSD as an HD cache / ZFS L2ARC (1)

h4rr4r (612664) | more than 4 years ago | (#33399926)

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.

Hey if Phoronix says it, it has to be true! (3, Funny)

Beelzebud (1361137) | more than 4 years ago | (#33399782)

I hear that every install of ZFS for Linux comes with a pre-installed Steam client, and a free copy of Team Fortress 2 For Linux!

Re:Hey if Phoronix says it, it has to be true! (1)

TheGratefulNet (143330) | more than 4 years ago | (#33399800)

...or the yahoo toolbar.

(why is oracle pimping the yahoo toolbar, btw, on their java installs? boggle!)

Re:Hey if Phoronix says it, it has to be true! (1)

binarylarry (1338699) | more than 4 years ago | (#33399836)

Sun started that BS.

Re:Hey if Phoronix says it, it has to be true! (2, Informative)

h4rr4r (612664) | more than 4 years ago | (#33399844)

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:Hey if Phoronix says it, it has to be true! (2, Interesting)

CoolGopher (142933) | more than 4 years ago | (#33400276)

I came across this job post [valvesoftware.com] at Valve just recently. To save you from having to follow the link, it includes the item
"Port Windows-based games to the Linux platform".

Just sayin'.

Re:Hey if Phoronix says it, it has to be true! (1)

Beelzebud (1361137) | more than 4 years ago | (#33400310)

You do realize that Valve has Linux server clients for all of their games, and the Steam infastructure itself runs on Linux right?

Just sayin'.

New Project (1)

bragr (1612015) | more than 4 years ago | (#33399812)

Looks like I'll finally have a reason to get around to rebuilding my home server. Well, as soon as WD finishes warrantying one of the drives in the array that is. Current estimates put ZFS and the drive arriving at the same time!

Linux has BTRFS! (-1)

Anonymous Coward | more than 4 years ago | (#33399870)

Didnt think ZFS mattered mush anymore with the introduction/inclusion of BTRFS in Linux.

ZFS recap (2, Funny)

oldhack (1037484) | more than 4 years ago | (#33399908)

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?

Re:ZFS recap (2, Funny)

colinrichardday (768814) | more than 4 years ago | (#33399960)

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:ZFS recap (1)

Miseph (979059) | more than 4 years ago | (#33400088)

Perhaps you'd like a colonel to read it for you?

Re:ZFS recap (3, Informative)

Anonymous Coward | more than 4 years ago | (#33400274)

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.youtube.com/watch?v=OdHUub462pM

Though I recommend you set aside the three hours (even if it's over several days) to really get a good understanding of how things work.

WHAT WOULD ELLISON DO ?? CALL THE CAV ?? (-1, Offtopic)

Anonymous Coward | more than 4 years ago | (#33399936)

The Indians need not fear the white man.

Lack of professionalism? (1, Interesting)

Anonymous Coward | more than 4 years ago | (#33399958)

This sounds great.

However, some of the authors comments in the Phoronix thread (such as "FUSE is crap") -- and the fact that he's announcing this on Phoronix instead of via some other technical channel temper my enthusiasm just a bit...

Re:Lack of professionalism? (1)

Beelzebud (1361137) | more than 4 years ago | (#33399998)

Yeah his comments on that forum are extremely childish. Too many "dudes" for me to even take seriously.

Useless without dedupe (1, Flamebait)

Zlurg (591611) | more than 4 years ago | (#33400188)

Just ask freenas and openfiler and, heck, even opensolaris
Oh, and this:
23212 [root@place]/mnt/Scratch: zfs get utf8only
bad property list: invalid property 'utf8only'

Know why? Cuz you don't get any choice in the matter outside Solaris proper.
Without dedupe and without the ability to use non-UTF8 in filenames (try creating a directory called Télépopmusik in a ZFS pool), ZFS is a future-former. I hear they name filesystems after murderers these days just for the street cred.
You did not read it here first, but I'll take the heat.

Can I remove a disk from it yet? (5, Interesting)

Daffy Duck (17350) | more than 4 years ago | (#33400210)

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?

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?