VSquared56 writes, "Novell announced a shift in the default filesystem from ReiserFS to ext3 for users of its SuSE Enterprise Linux. This news comes shortly after Hans Reiser's arrest, though Novell says the decision was being considered long before. Though Novell will continue supporting ReiserFS 3, it claims ext3 is more stable and will 'soon' match performance with the newer ReiserFS 4. What implications will this have for SuSE users, and ReiserFS's future as a whole?"
Did Novell ever get around to porting Novell Storage Services [NSS] to Linux?
NSS was the B-Tree successor to the old allocation table NetWare file system, and it had all the permissions and attributes that were unique to the Novell World:
Read
Write
Create
Erase
Modify
File Scan
Access Control
Supervisor
So did Novell ever get around to porting an R/W/C/E/M/FS/AC/S file system to Linux, to be used in place of the standard Unix RWX/RWX/RWX file system?
I think is to have a poll as to measure people's opinions about the guy's innocence. With options such as 1. He is innocent 2. He is guilty 3. Cowboyneal did it etc..
7. "I'm just a patsy-OH NO I'VE BEEN SHOT!" (Oswald defence) 8. "You want the Truth? YOU CAN'T HANDLE THE TRUTH!!" (A Few Good Men defence) 9. "I'm telling you, it wasn't me! It was a one-armed man! You've got to believe me!" (The Fugitive defence) 10. "These are not the evidence you are lookiing for" (Obi-wan defense) 11. "That depends on what the definition of 'kill' is." (The Clinton defence) 12. "Putting this 'evidence' out for anyone to read is helping our terrorist enemies." (The Bush defence)
I think anyone capable of writing a filesystem is capable of killing a person. But also anyone writing an open source software is too kind to kill a person.
why not move to xfs? it's a very good performance file system.
XFS is high performance especially for large files and multitasked access. reiserfs (3) is high performance especially for small files and singletasked access. JFS is also a good journalled file system with many nifty features, although perhaps not as mature as XFS.
unless there a rumours of the author being a murderer of course.
Neither X nor J have been accused of murder, to my knowledge.
There have been too many reports in the last couple of months of people whose machines have lost power, and booted up, only to find that every file on their XFS filesystems has been filled with zeroes.
For the record, I was also quite underwhelmed by XFS. The Gentoo people, I think, wrote that XFS is primarily to large files and *only* if you have an UPS (and proper shutdown control). The problem is that it (quite aggressively) cache write-data; I have seen data disappear which was written nearly 2 hours before.
I am quite happy with ext3. Reiserfs had a nasty tendency to slowly deteriorate over time, becoming slower and slower.
There have been too many reports in the last couple of months of people whose machines have lost power, and booted up, only to find that every file on their XFS filesystems has been filled with zeroes. That's what backups are for. Seriously, with XFS you run a very real risk of zeroing out a file if the file system isn't shut down properly. But with reiserfs, you run a very real risk of losing the file system. In over a decade, I've never seen that happen with XFS -- only zeroed out files.
Actually, the zero-filled files is a misfeature of XFS. Having a UPS will not save you. There are two XFS problems:
1. Power loss can destroy your filesystem. Solution: do not use XFS or ReiserFS without a UPS.
2. An unclean shutdown can leave you with zero-filled files. AFAIK this is a design flaw in XFS or, depending how you look at it, a tradeoff of data integrity for performance. If you don't like the tradeoff then your only choice is to use another filesystem.
What good is a UPS going to do in the case the machine powers off because of a problem with the power unit, a motherboard short circuit, and so on? Any filesystem with serious data loss on a power failure is not acceptable, period.
No, this happens because it's the way XFS does journalling.
XFS journalling isn't as good as the one in ext3, from users' POV. Ext3 default journaling mode takes care of the relationship between metadata and the data associated to that metadata (and here let me remember that journalling/softupdates is a way to avoid corruption of the *metadata*, if you lose data because of a power cut that's fine, but it's not fine that the filesystem gets damaged and needs fsck because the metadata got corrupted)
IOW: when ext3 is going to write metadata to the disk, it looks first to the dirty data cached in the memory and writtes the data *before* it writes the metadata.
XFS journaling, in the other hand, does *not* care about writing the data before the metadata. Why? Well, because journalling is about keeping the metadata safe so you don't need fsck. This means that in case of a power cut, XFS may leave the contents of a "file" (metadata) unscycrhonized with its data. Because of that, the metadata may be pointing to random free zone of the disc with confidential information (passwords) which was deleted but it has not been overwritten, so XFS sets it to zero for safety. Ext3, on the other hand, will never left your data "unscychornised" with your metadata. The file may get corrupted because the program that was manipulating it was stopped in the power cut, but the relationship between the data and the metadata is always coherent.
Ext3 journaling mode may be considered an "extra", and it *does* pay a performance disadvantage because of this. If you want ext3 to behave like xfs (and get better performance), mount your fs with the mount option "data=writeback". Reiserfs in the other hand historically had a similar journaling method as XFS (just like JFS), but the suse guys created a journaling mode similar to the default one in ext3 which AFAIK is not enabled by default (at least on mainline) and gets enabled with "data=ordered"
Is the XFS journaling mode worse? Well, for desktop users, who would rather have syncronized their data and their metadata, clearly yes. This is why XFS is just not the best FS for desktops - its a wonderful FS, but just not "optimized" for desktops. NTFS journaling does the same that ext3 does, BTW, and it's for a reason.
It's interesting that you mention that. Some time ago, I used ReiserFS as the filesystem on my laptop computer (I only have one partition, not counting swap). The performance was alright and it always took some seconds to mount the partition (this is a known thing for ReiserFS). So, more or less, my experience had been fine. One day, I was trying to view one JPG file and the program was unable to open it, so I wondered why. After examining the file, I found out that while the file size was alright, its contents were all binary zeros. I discovered similar things for a handful of files in my system, many of them in my home directory, I supposed because that's where the biggest part of the disk data is located and if a problem arose, it's probably going to be there.
At the beginning I suspected something had gone wrong while copying the data to an external USB hard drive and back to the newly formatted ReiserFS partition. But, some weeks later, I discovered a similar situation in a file I had created recently (after the data move), and that had been available there for many days. I am only a desktop user and I lack evidence on what caused this, but I tested my harddrive to see if it had bad sectors or behaved poorly for some reason, and nothing turned up. I fsck'ed the partition and everything was alright. I suspected this problem was due to ReiserFS, so I took the decision of switching back to ext3 with dir_index activated, and the problem hasn't reappeared again. I suspect I hit a bug in the ReiserFS code, and I lost my data in one or several of those ocasions when I left my laptop alone for some time and it powered off suddenly when it ran out of battery. This happened more times since the switch to ext3, but I haven't lost any more files since then.
I know this can be a particular case which may not represent the behaviour of ReiserFS, but as I read your comment I thought I had to share my experience too.
Yup, no surprise there. XFS caches writes very agressively in ram, around 50MB if not more, for long periods of times. So it "feels" fast but really isn't in some aspects.
So you pop the power off and *wham* bye bye cached data. This is definately not any kind of fun.
XFS was written for environments where the power just dooes not go out -- datacenters, people with a very good UPS etc. I generally recommend XFS for people with lots of large files, but if they don't have a good power backup, I change my mind.
Neither xfs nor jfs partitions can be reduced. This may not be a big deal to compaines who just add disks and expand their partitions, but I know that I lost about two hundred gig worth of data that would probably still be around if I could have reduced my jfs partition. After that I tried to install ReiserFS then Reiser4, and after a little bit of trouble with those, decided I'd use Ext3 because it just works. Even if its performance isn't as great as some other file systems, I don't know too many people who have lost data because of flaws (or "features") of the filesystem.
At least that's what happens to a sinking ship. A maintainer going missing does not quite instill the users
with confidence, especially when it is happening due to reasons other than flagging interest. Most commercial
distributions have SLAs which sort of work against such brilliant work by an individual contributor - they
just can't depend on the whims of a person or his fate.
One of my friends once told me that "Extraordinary hackers are people with socially acceptable problems".
In fact to achieve what they feel they must, a lot of them give up a lot - health, social lives and financial security. But because a few do that, does not mean FOSS programmers are crackpots [gmane.org].
And I say this as a son who's home (which I can because my commits go to a public CVS)
watching over a sick father.
So as understandable as it is that commercial vendors might want to switch away, but that doesn't mean
anyone gets to shine a torch or make jokes [reboot.net.au] into somebody else's darkness.
Almost right, but the fact is Hans Reiser wasn't reisersfs3 maintainer. He long ago declared version 3 was dead and only reiser4fs worth using. reiserfs3 was maintained mainly by one guy in SUSE, who became fed up with it. And rightly recomended going to ext3/ext4.
Just BTW, I am using reiserfs3 on my system and I thinking about migration to some FS with future.
This news comes shortly after Hans Reiser's arrest
That news was this week. This news from SuSE, however, is very old [wordpress.com] already and apparently they indeed decided about this before Reiser got arrested.
It's also interesting how people now explain the blood on Reiser's shirt in this comic [geekz.co.uk], while this comic also predates this whole arrest story.:-)
It seems that ReiserFS really depends on 1 guy. For any company this is a risk. It sounds reasonable to me to stay away from products and features like that.
Hans Reiser really hasn't been supporting Reiserfs (v3) for a while now, most of the work has been coming out of SuSE anyway. At least, the patches we've been running (mostly merged into mainline now) have been coming from them, and we're not even a customer.
I guess we won't be bugging Hans about issues for a bit now. Hope he's innocent and it gets resolved quickly - must suck for everyone involved right now.
Several commenters appear to think that this is due to the arrest of Hans, In fact it was announced over a month ago, before any of the stories about Hans broke. The original announcement is from the 14th september http://lists.opensuse.org/opensuse-factory/2006-09 /msg00542.html [opensuse.org]
by Anonymous Coward
on Sunday October 15 2006, @07:05AM (#16443005)
When it comes to performance between the filesystems (reiser vs. ext3 vs. xfs) then I don't have much to comment, but with regards to security... I've used reiser for quite some time but in the end threw it away because it just couldn't cope with what I wanted..
First your average backup. Yes, I'm well aware that you can always tools like tar but really.. Its the same deal with Sun's current development ZFS: it lacks the option to decently make a backup. Yes you can use tar, but I don't consider this decent. I'm talking about tools like backup/restore (ext3) or even native "ports" like xfsdump/xfsrestore. Easy, fast and reliable. Make a whole dump (or increamental), you can then either restore the whole session or use an interactive shell to merely grab the file(s) you're after. Naturally it also supports commandline parameters. And Reiser? IIRC (correct me if I'm wrong please) its even longer around than xfs, and even xfs managed to get me something decent for making backups...
Last but not least; crash recovery. I know, this is threading on thin ice since these results cannot be reproduced perse but the whole nature of reiser makes it good and bad for workstations (like SuSE). The good part is its speed, the way it caches and writes data in such a way where it tries to store things in one specific part makes it faster. I can't comment if reiser really is faster than others, I never noticed it. But the bad part is also that if you have a crash on your hands (just turn of your computer right now. No, not a shutdown but keep the powerbutton pressed untill it goes "poof") and reboot chances are very high that you just lost valuable data.
The theory behind journaling should give you some protection against this, and normally it does, but its my experience that whenever something like this happened on a box which was using reiser I lost just too many files. Several files in/etc used to become corrupt, binaries started going haywire and the worst part: because the index wasn't affected it was quite hard to detect these bad files.
Eventually I moved to XFS myself and never bothered looking back. Its not perfect, absolutely not since on XFS you too can experience situations like I just described. But in that same environment where I sometimes had to endure a powerloss I noticed that the frequency in which my data became corrupt was far and far less than with reiser. So my conclusion: reiser isn't the best when it comes to keeping your data safe. Its also a conclusion which has been backed up by other people who experiences the same problems in a more or lesser degree.
So my comment: finally Novell is coming to its senses. IMO they should have done this years ago, either going to XFS (my favorite) or ext3 where the latter is ofcourse the most logical choice considering how this evolved from ext2 (which, strangely enough, used to be the default on SuSE. I never did understand why they'd move away from it).
Let me start by saying I'm not disagreeing with you, just pointing out that different people may have different experiences (AKA YMMV).
``First your average backup. Yes, I'm well aware that you can always tools like tar but really.. Its the same deal with Sun's current development ZFS: it lacks the option to decently make a backup. Yes you can use tar, but I don't consider this decent. I'm talking about tools like backup/restore (ext3) or even native "ports" like xfsdump/xfsrestore. Easy, fast and reliable. Make a whole dump (or increamental), you can then either restore the whole session or use an interactive shell to merely grab the file(s) you're after. Naturally it also supports commandline parameters. And Reiser? IIRC (correct me if I'm wrong please) its even longer around than xfs, and even xfs managed to get me something decent for making backups...''
I believe backup tools that depend on the specifics of filesystems are a bad idea.
When you go looking for filesystem-independent backup tools, I'm sure you'll find plenty (the recent thread here on Slashdot may be a good starting point). I myself keep most of my data in Subversion repositories and databases; backups are made through the appropriate backup tools. Whatever is left on the filesystem is synchronized between a couple of computers using rsync.
``The good part is its speed, the way it caches and writes data in such a way where it tries to store things in one specific part makes it faster. I can't comment if reiser really is faster than others, I never noticed it.''
In the tests I ran, it wiped the floor with ext2 and (OpenBSD) ffs, especially when extracting lots of small files. I have no idea how it compares to more modern filesystems like XFS, ZFS, etc.
``But the bad part is also that if you have a crash on your hands (just turn of your computer right now. No, not a shutdown but keep the powerbutton pressed untill it goes "poof") and reboot chances are very high that you just lost valuable data.''
Although I have lost files on ReiserFS partitions, I've lost way more on ext2 and (especially) HFS+ partitions.
``The theory behind journaling should give you some protection against this, and normally it does, but its my experience that whenever something like this happened on a box which was using reiser I lost just too many files. Several files in/etc used to become corrupt, binaries started going haywire and the worst part: because the index wasn't affected it was quite hard to detect these bad files.''
Often when files seem to be missing after a crash, fsck has been able to recover them for me. This goes for ext2, reiserfs, ffs, and hfs+. Reiserfs is the only one of these on which I have never gotten the filesystem so broken it couldn't be fixed anymore.
In case people are wondering where I get my data from: I work with a lot of old hardware which sometimes fails, laptops that run out of battery or are dropped on the floor, accidentally unplugged power cables, and the occasional unclean shutdown.
Geez, now blood's found in his car, and with the passenger seat missing [mercurynews.com], history of abuse, guy is arrested with $8,900 and his passport on him...
If he were a famous football player, he'd have a chance, but I don't think a filesystem developer can muster up a "dream team".
it claims ext3 is more stable and will 'soon' match performance with the newer ReiserFS 4.
Gee, ext3 must've matured a lot in the past few years. I stopped using extX filesystems long ago because they lost files after power cuts waay too easily. ( I could bork an old RedHat install simply by pulling the plug/rebooting several times ). Moved to reiser then xfs and barely lost anything if I had to force a reboot.
I stopped using extX filesystems long ago because they lost files after power cuts waay too easily.
That's still better than reiserfs, which does not need a power cut in order to lose data. I still recall a comment from a tech support area I used to frequent: "reiserfs runs really fast until it crashes and you lose all your data. As a result it has a lot of ex-users who are now sadder but wiser."
It is also important to remember that ext3 can be configured for a number of different points along the speed/safety tradeoff, so any stories about problems (with speed *or* safety) should state which mode they were using.
Perhaps. But the single most valuable thing about both Reiser filesystems is how well they handle large numbers of small files. I hate Berkeley DB and its ilk with a passion. They take all kinds of valuable data that should be addressable with standard tools and obscure it in some weird format that I can't make any sense of without some specialized set of tools. Not only that, but they're slow!
I want to stop using these awful things. I want to use a hierarchical naming scheme to address the individual bits of data I'm stuffing into the filesystem without having to resort to stupid tricks with splitting up the name so I don't have anymore than 256 entries per directory.
None of the filesystems made for Linux aside from ReiserFS seem to even acknolwedge that this problem is worth solving. Personally, I think it is a major, short-sighted bellybutton gazing failure. The excuse seems to be "Well, you're using the filesystem in a strange way that nobody uses it in, so stop doing that!". But that's a completely circular argument. I simply do not WANT to contort my programs in such a ridiculous way to accomodate the failings of filesystem designs.
ReiserFS is fast and flexible. I've never had any data loss with 3. At least, not in the last 3 years or so. And I have a machine that will (for reasons of a bad motherboard) randomly lock up if I'm using both the disk and the Ethernet card heavily.
I don't really care that much about plug-ins. They're kind of a neat idea for having super-efficient storage for caches and stuff, but really I just want to be able to independently address millions of small pieces of data and have it be reasonably efficient.
Editors, if you're the ones doing it, please stop. If submitters are doing it, please edit their submissions. We don't need this Roland Piquipaille/Ric Romero style of foolishness, i.e. "Blah blah has happened to company FOO, what do you all think?" Posting it for discussion on Slashdot IMPLIES you're going to get a million different viewpoints, none of which are really important to the submitter. You'll get the viewpoints anyway, you don't need to "prompt" us for them.
I wrote the original email proposing that SUSE switch from reiserfs to ext3. At the risk of triggering responses of "The lady doth protest too much," I'll restate a few statements I've made elsewhere in response to common questions:
1) The decision has *nothing* to do with Hans' situation. The email was released on the same day as the initial story broke, but it was pointed out to me after I had sent the email. I was concerned then, correctly as it turns out, that people would consider the two issues intertwined. They're not. My proposal was based on technical and maintainability reasons alone. The timing is an extremely unfortunate coincidence.
2) SUSE is *not* dropping reiser3 support. This change only affects the default. It doesn't change our support of reiser3 at all. We still support four major file systems: ext3, reiserfs, xfs, and ocfs2. Our installer offers other file systems as well as a convenience, and users are free to use any of them. So, if you're committed to reiser3 or xfs, nothing is stopping you from continuing to deploy systems using them.
3) Many benchmarks show reiser3 as performing better than ext3, and this is mostly true. What isn't shown in those benchmarks is that if you're operating two or more reiser3 file systems in parallel, performance will degrade for both of them due to the use of BKL everywhere. ext3 (and other file systems) will don't degrade in that case. I've also read reports that there is a bit of research going on into making ext3 locking finer grained. I don't have any sources to cite, but any reduction of critical sections without reducing reliability is always a good thing.
People refer to reiser3 as a modern file system, but I'd call it progressive. Reiser3 has served us well for years, but it's showing its age. The basic idea behind reiser3 is still sound, and when extended with integrated integrity checking and better b-tree locking borrowed from years of database research, it would perform extremely well. The problem is that adding the first is a huge disk format change, which means it's no longer reiser3. Adding the second is a hugely invasive change that would throw out a good chunk of the existing code -- again, essentially creating a new file system. It would be like people saying, "I like my ext3 file system, but I don't like the code. Let's start over." Combined with a small development community, it's a recipe for instability and there are more interesting problems out there.
Ext3, in contrast, is stable and likely will match ReiserFS's performance advantages "soon."
Rubbish. Ext3 has never been able to match Reiser's performance on small files or in other areas, and the notion that ext3 is going to match it is absurd. Even ext4 is not likely to catch up. A lot of ext developers have bizarre ideas about how their filesystem compares to Reiser, XFS or even JFS in a lot of areas. Ext is simply a stable and solid, but badly evolved, filesystem and it is a filesystem that generates an awful lot of disk activity.
Late last year I was researching Reiserfs and Ext3 to see which would be best suited for my new server.
Resierfs looked like the clear winner for two good reasons:
1. Reiserfs is faster. Much faster than ext3 in nearly every scenario. Large files and small files. 2. No inode problems. If your users fill your HD with hundreds of thousands of tiny files you're not going to run out of inodes before you run out of disk space. This is something that needs to be anticipated (at the cost of more disk space) at filesystem creation time in ext3.
Reliability for both filesystems was pretty much the same from all accounts.
But in the end I went with ext3 for one and only one reason: Recoverability.
Reiserfs had no, or very few decent, recovery utilities. If a filesystem corruption occurred (and it seemed that the probablity of such corruptions was equal for both filesystems), then data on an ext3 fs stands a much better chance of being recovered than on a reiserfs one.
Of course that was late 2005; that situation may have changed by now.
I predicted this a few days ago, when I wrote here:
That it/is/ going to damage reiserfs is beyond any doubt, no matter whether he's proven innocent, not proven guilty, or proven guilty. The name is tainted, and a business executive will not likely touch anything related to that person, no matter whether it gets taken over and run by other people or not.
This was modded flamebait.
People, you might not want to hear it, and you might not agree with stupid knee-jerk reactions, but these reactions will be coming. The name "reiserfs" is tainted, whether that's rational or not.
Unfortunately, I hate to agree with you, but it's true. If they project renames and then it continues, it might get picked up again.
The other concern is going to be about support, if Hans is found guilty or not, it doesn't really matter. A company such as Novell may consider that the filesystem platform isn't as supported as what it once was and is moving away from it.
From a marketing point of view, Novell won't want to associated with it either. If they show support for him, and he is found guilty, it's a marketing nightmare for Novell.
From a marketing point of view, Novell won't want to associated with it either. If they show support for him, and he is found guilty, it's a marketing nightmare for Novell.
Are you kidding - this stuff practically writes itself.
"Novell and ReiserFS - we have THE killer linux filesystem!"
"Novell and ReiserFS - your files are as secure as the county lockup."
"Novell and ReiserFS - while the jury's out, we continue give you a choice of file systems"
"Novell and ReiserFS - exabytes of data - and only one piece of information missing..."
"Novell and ReiserFS - no dead inodes, no dead files, no dead bodies..."
I've been to the namesys [namesys.com] website and haven't found anything about the Hans Reiser arrest. If it is more than a one man show, then they should have a prominent statement about their intention to continue development regardless of the outcome. Not seeing something to that effect after this mych time would make me quite nervous if I had a business or product that relied on continued development of the ReiserFS line. Seems as if Namesys is accepting the inevitable demise of the whole organization at this point; I hope that changes.
I hope he is not guilty; I hope that his wife simply tried to pull a scam and is actually OK and is found hiding in Russia or somewhere. Even though it'd have been a scummy thing for her to do, at least their kids would not have lost their mother. That's the single worst part of the whole thing -- oh wait, am I saying "think of the children?" Shoot me now, please!
Not that I've got my finger right on the pulse of FS development, but I find it hard to believe that ext3 is soon going to equal Reiserfs for all cases. Perhaps for a typical case, but ReiserFS was supposed to allow a lot of stuff that was not feasible with ext3 like efficiently having really small files, using the FS as a database, and a lot of other potentially groundbreaking research and abilities. I hope none of the good ideas get lost.
What would happen if (god forbid) Linus were in Reiser's place? Would everyone here be distancing themselves from using Linux? Would Novell, IBM, etc. abandon the use of Linux, throwing out the baby with the bathwater?
This is why when the story hit I posed legitimate questions regarding the filesystem's future (and got flamed for it, BTW, here and on linuxquestions); a person's career work should be viewed independently of his or her personal misdeeds. Otherwise, we should abandon electricity and incandescent lights (Edison was a bit of a bastard, and his invention of the electric chair "tainted" AC), jets (Heinkel was a nazi), Mercury and Apollo programs should never have happened (Wernher von Braun, the brain behind those programs, was a nazi, willing or otherwise). There are many, many worthwhile inventions proposed, designed, and/or implemented by evil people, and yet we use them on a daily basis, because regardless of the creators' nature, philosophy, or misdeeds, they have produced some worthwhile things that abandoning them because of the heritage would be somewhere between silly and irresponsible.
Hurm (Score:5, Funny)
Well, just a guess . . . but they might have to use a new filesystem!
No one laughed last time... (Score:5, Funny)
Parent
Novell Storage Services [NSS] R/WC/E/M/FS/AC/S? (Score:4, Interesting)
Did Novell ever get around to porting Novell Storage Services [NSS] to Linux?
NSS was the B-Tree successor to the old allocation table NetWare file system, and it had all the permissions and attributes that were unique to the Novell World:
So did Novell ever get around to porting an R/W/C/E/M/FS/AC/S file system to Linux, to be used in place of the standard Unix RWX/RWX/RWX file system?
And if so, is anyone out there using it?
Parent
Have you ever heard the phrase: (Score:3, Funny)
The Slashdot Way... (Score:5, Funny)
about the guy's innocence. With options such as
1. He is innocent
2. He is guilty
3. Cowboyneal did it etc..
You're forgetting (Score:5, Funny)
5. The glove's too tight (OJ)
6. Is that Chewbacca here? (Chewbacca defense)
Parent
Re:You're forgetting (Score:5, Funny)
8. "You want the Truth? YOU CAN'T HANDLE THE TRUTH!!" (A Few Good Men defence)
9. "I'm telling you, it wasn't me! It was a one-armed man! You've got to believe me!" (The Fugitive defence)
10. "These are not the evidence you are lookiing for" (Obi-wan defense)
11. "That depends on what the definition of 'kill' is." (The Clinton defence)
12. "Putting this 'evidence' out for anyone to read is helping our terrorist enemies." (The Bush defence)
Parent
Re:The Slashdot Way... (Score:4, Funny)
Parent
xfs for ever (Score:3, Interesting)
Re:xfs for ever (Score:5, Funny)
XFS is high performance especially for large files and multitasked access.
reiserfs (3) is high performance especially for small files and singletasked access.
JFS is also a good journalled file system with many nifty features, although perhaps not as mature as XFS.
Neither X nor J have been accused of murder, to my knowledge.
All hail J.
Regards,
--
*Art
Parent
Re:xfs for ever (Score:5, Interesting)
Parent
Re:xfs for ever (Score:4, Interesting)
Parent
Re:xfs for ever (Score:4, Funny)
If you don't know what you are doing, you should stick with the default "distro".
And Microsoft recommends NTFS.
Parent
Re:xfs for ever (Score:5, Funny)
Parent
Re: (Score:3, Interesting)
That's what backups are for. Seriously, with XFS you run a very real risk of zeroing out a file if the file system isn't shut down properly. But with reiserfs, you run a very real risk of losing the file system. In over a decade, I've never seen that happen with XFS -- only zeroed out files.
Regards,
--
*Art
Re: (Score:3, Interesting)
1. Power loss can destroy your filesystem. Solution: do not use XFS or ReiserFS without a UPS.
2. An unclean shutdown can leave you with zero-filled files. AFAIK this is a design flaw in XFS or, depending how you look at it, a tradeoff of data integrity for performance. If you don't like the tradeoff then your only choice is to use another filesystem.
Source: http://linuxmafia.com [linuxmafia.com]
Re:xfs for ever (Score:5, Insightful)
Parent
Re:xfs for ever (Score:5, Informative)
No, this happens because it's the way XFS does journalling.
XFS journalling isn't as good as the one in ext3, from users' POV. Ext3 default journaling mode takes care of the relationship between metadata and the data associated to that metadata (and here let me remember that journalling/softupdates is a way to avoid corruption of the *metadata*, if you lose data because of a power cut that's fine, but it's not fine that the filesystem gets damaged and needs fsck because the metadata got corrupted)
IOW: when ext3 is going to write metadata to the disk, it looks first to the dirty data cached in the memory and writtes the data *before* it writes the metadata.
XFS journaling, in the other hand, does *not* care about writing the data before the metadata. Why? Well, because journalling is about keeping the metadata safe so you don't need fsck. This means that in case of a power cut, XFS may leave the contents of a "file" (metadata) unscycrhonized with its data. Because of that, the metadata may be pointing to random free zone of the disc with confidential information (passwords) which was deleted but it has not been overwritten, so XFS sets it to zero for safety. Ext3, on the other hand, will never left your data "unscychornised" with your metadata. The file may get corrupted because the program that was manipulating it was stopped in the power cut, but the relationship between the data and the metadata is always coherent.
Ext3 journaling mode may be considered an "extra", and it *does* pay a performance disadvantage because of this. If you want ext3 to behave like xfs (and get better performance), mount your fs with the mount option "data=writeback". Reiserfs in the other hand historically had a similar journaling method as XFS (just like JFS), but the suse guys created a journaling mode similar to the default one in ext3 which AFAIK is not enabled by default (at least on mainline) and gets enabled with "data=ordered"
Is the XFS journaling mode worse? Well, for desktop users, who would rather have syncronized their data and their metadata, clearly yes. This is why XFS is just not the best FS for desktops - its a wonderful FS, but just not "optimized" for desktops. NTFS journaling does the same that ext3 does, BTW, and it's for a reason.
Parent
Re:xfs for ever (Score:5, Insightful)
OMG, are you kidding? If it was NTFS or FAT, people on
Parent
Re:xfs for ever (Score:5, Interesting)
At the beginning I suspected something had gone wrong while copying the data to an external USB hard drive and back to the newly formatted ReiserFS partition. But, some weeks later, I discovered a similar situation in a file I had created recently (after the data move), and that had been available there for many days. I am only a desktop user and I lack evidence on what caused this, but I tested my harddrive to see if it had bad sectors or behaved poorly for some reason, and nothing turned up. I fsck'ed the partition and everything was alright. I suspected this problem was due to ReiserFS, so I took the decision of switching back to ext3 with dir_index activated, and the problem hasn't reappeared again. I suspect I hit a bug in the ReiserFS code, and I lost my data in one or several of those ocasions when I left my laptop alone for some time and it powered off suddenly when it ran out of battery. This happened more times since the switch to ext3, but I haven't lost any more files since then.
I know this can be a particular case which may not represent the behaviour of ReiserFS, but as I read your comment I thought I had to share my experience too.
Parent
Re:xfs for ever (Score:4, Informative)
So you pop the power off and *wham* bye bye cached data. This is definately not any kind of fun.
XFS was written for environments where the power just dooes not go out -- datacenters, people with a very good UPS etc. I generally recommend XFS for people with lots of large files, but if they don't have a good power backup, I change my mind.
Parent
Re:xfs for ever (Score:4, Interesting)
Parent
Rats first and Captain last (Score:4, Interesting)
At least that's what happens to a sinking ship. A maintainer going missing does not quite instill the users with confidence, especially when it is happening due to reasons other than flagging interest. Most commercial distributions have SLAs which sort of work against such brilliant work by an individual contributor - they just can't depend on the whims of a person or his fate.
One of my friends once told me that "Extraordinary hackers are people with socially acceptable problems". In fact to achieve what they feel they must, a lot of them give up a lot - health, social lives and financial security. But because a few do that, does not mean FOSS programmers are crackpots [gmane.org]. And I say this as a son who's home (which I can because my commits go to a public CVS) watching over a sick father.
So as understandable as it is that commercial vendors might want to switch away, but that doesn't mean anyone gets to shine a torch or make jokes [reboot.net.au] into somebody else's darkness.
Re:Rats first and Captain last (Score:5, Informative)
Just BTW, I am using reiserfs3 on my system and I thinking about migration to some FS with future.
Parent
Old news (Score:5, Informative)
It's also interesting how people now explain the blood on Reiser's shirt in this comic [geekz.co.uk], while this comic also predates this whole arrest story.
arrest aside... (Score:4, Insightful)
Re: (Score:3)
I guess we won't be bugging Hans about issues for a bit now. Hope he's innocent and it gets resolved quickly - must suck for everyone involved right now.
Nothing todo with Hans' arrest. (Score:5, Informative)
Smart move, just a little late (Score:5, Informative)
First your average backup. Yes, I'm well aware that you can always tools like tar but really.. Its the same deal with Sun's current development ZFS: it lacks the option to decently make a backup. Yes you can use tar, but I don't consider this decent. I'm talking about tools like backup/restore (ext3) or even native "ports" like xfsdump/xfsrestore. Easy, fast and reliable. Make a whole dump (or increamental), you can then either restore the whole session or use an interactive shell to merely grab the file(s) you're after. Naturally it also supports commandline parameters. And Reiser? IIRC (correct me if I'm wrong please) its even longer around than xfs, and even xfs managed to get me something decent for making backups...
Last but not least; crash recovery. I know, this is threading on thin ice since these results cannot be reproduced perse but the whole nature of reiser makes it good and bad for workstations (like SuSE). The good part is its speed, the way it caches and writes data in such a way where it tries to store things in one specific part makes it faster. I can't comment if reiser really is faster than others, I never noticed it. But the bad part is also that if you have a crash on your hands (just turn of your computer right now. No, not a shutdown but keep the powerbutton pressed untill it goes "poof") and reboot chances are very high that you just lost valuable data.
The theory behind journaling should give you some protection against this, and normally it does, but its my experience that whenever something like this happened on a box which was using reiser I lost just too many files. Several files in
Eventually I moved to XFS myself and never bothered looking back. Its not perfect, absolutely not since on XFS you too can experience situations like I just described. But in that same environment where I sometimes had to endure a powerloss I noticed that the frequency in which my data became corrupt was far and far less than with reiser. So my conclusion: reiser isn't the best when it comes to keeping your data safe. Its also a conclusion which has been backed up by other people who experiences the same problems in a more or lesser degree.
So my comment: finally Novell is coming to its senses. IMO they should have done this years ago, either going to XFS (my favorite) or ext3 where the latter is ofcourse the most logical choice considering how this evolved from ext2 (which, strangely enough, used to be the default on SuSE. I never did understand why they'd move away from it).
Re:Smart move, just a little late (Score:4, Insightful)
``First your average backup. Yes, I'm well aware that you can always tools like tar but really.. Its the same deal with Sun's current development ZFS: it lacks the option to decently make a backup. Yes you can use tar, but I don't consider this decent. I'm talking about tools like backup/restore (ext3) or even native "ports" like xfsdump/xfsrestore. Easy, fast and reliable. Make a whole dump (or increamental), you can then either restore the whole session or use an interactive shell to merely grab the file(s) you're after. Naturally it also supports commandline parameters. And Reiser? IIRC (correct me if I'm wrong please) its even longer around than xfs, and even xfs managed to get me something decent for making backups...''
I believe backup tools that depend on the specifics of filesystems are a bad idea.
When you go looking for filesystem-independent backup tools, I'm sure you'll find plenty (the recent thread here on Slashdot may be a good starting point). I myself keep most of my data in Subversion repositories and databases; backups are made through the appropriate backup tools. Whatever is left on the filesystem is synchronized between a couple of computers using rsync.
``The good part is its speed, the way it caches and writes data in such a way where it tries to store things in one specific part makes it faster. I can't comment if reiser really is faster than others, I never noticed it.''
In the tests I ran, it wiped the floor with ext2 and (OpenBSD) ffs, especially when extracting lots of small files. I have no idea how it compares to more modern filesystems like XFS, ZFS, etc.
``But the bad part is also that if you have a crash on your hands (just turn of your computer right now. No, not a shutdown but keep the powerbutton pressed untill it goes "poof") and reboot chances are very high that you just lost valuable data.''
Although I have lost files on ReiserFS partitions, I've lost way more on ext2 and (especially) HFS+ partitions.
``The theory behind journaling should give you some protection against this, and normally it does, but its my experience that whenever something like this happened on a box which was using reiser I lost just too many files. Several files in
Often when files seem to be missing after a crash, fsck has been able to recover them for me. This goes for ext2, reiserfs, ffs, and hfs+. Reiserfs is the only one of these on which I have never gotten the filesystem so broken it couldn't be fixed anymore.
In case people are wondering where I get my data from: I work with a lot of old hardware which sometimes fails, laptops that run out of battery or are dropped on the floor, accidentally unplugged power cables, and the occasional unclean shutdown.
Parent
RTFM (Score:4, Informative)
See Solaris ZFS Administration Guide, Chapter 6 Working With ZFS Snapshots and Clones [sun.com].
Parent
Thins aren't looking up for Hans. (Score:5, Informative)
If he were a famous football player, he'd have a chance, but I don't think a filesystem developer can muster up a "dream team".
I expect other distros will knee-jerk too.
$ mount
ext3 more reliable? Whatthe! (Score:5, Interesting)
Gee, ext3 must've matured a lot in the past few years. I stopped using extX filesystems long ago because they lost files after power cuts waay too easily. ( I could bork an old RedHat install simply by pulling the plug/rebooting several times ). Moved to reiser then xfs and barely lost anything if I had to force a reboot.
Re:ext3 more reliable? Whatthe! (Score:5, Interesting)
That's still better than reiserfs, which does not need a power cut in order to lose data. I still recall a comment from a tech support area I used to frequent: "reiserfs runs really fast until it crashes and you lose all your data. As a result it has a lot of ex-users who are now sadder but wiser."
It is also important to remember that ext3 can be configured for a number of different points along the speed/safety tradeoff, so any stories about problems (with speed *or* safety) should state which mode they were using.
Parent
ext3 Performance Matches Reiser4?! (Score:5, Insightful)
Huh? In whose benchmarks? What about space usage? What about plugins for arbitrary attributes?
Re:ext3 Performance Matches Reiser4?! (Score:5, Interesting)
Perhaps. But the single most valuable thing about both Reiser filesystems is how well they handle large numbers of small files. I hate Berkeley DB and its ilk with a passion. They take all kinds of valuable data that should be addressable with standard tools and obscure it in some weird format that I can't make any sense of without some specialized set of tools. Not only that, but they're slow!
I want to stop using these awful things. I want to use a hierarchical naming scheme to address the individual bits of data I'm stuffing into the filesystem without having to resort to stupid tricks with splitting up the name so I don't have anymore than 256 entries per directory.
None of the filesystems made for Linux aside from ReiserFS seem to even acknolwedge that this problem is worth solving. Personally, I think it is a major, short-sighted bellybutton gazing failure. The excuse seems to be "Well, you're using the filesystem in a strange way that nobody uses it in, so stop doing that!". But that's a completely circular argument. I simply do not WANT to contort my programs in such a ridiculous way to accomodate the failings of filesystem designs.
ReiserFS is fast and flexible. I've never had any data loss with 3. At least, not in the last 3 years or so. And I have a machine that will (for reasons of a bad motherboard) randomly lock up if I'm using both the disk and the Ethernet card heavily.
I don't really care that much about plug-ins. They're kind of a neat idea for having super-efficient storage for caches and stuff, but really I just want to be able to independently address millions of small pieces of data and have it be reasonably efficient.
Parent
Ending submissions with an idiotic question (Score:5, Insightful)
From the Horse's Mouth (Score:5, Informative)
1) The decision has *nothing* to do with Hans' situation. The email was released on the same day as the initial story broke, but it was pointed out to me after I had sent the email. I was concerned then, correctly as it turns out, that people would consider the two issues intertwined. They're not. My proposal was based on technical and maintainability reasons alone. The timing is an extremely unfortunate coincidence.
2) SUSE is *not* dropping reiser3 support. This change only affects the default. It doesn't change our support of reiser3 at all. We still support four major file systems: ext3, reiserfs, xfs, and ocfs2. Our installer offers other file systems as well as a convenience, and users are free to use any of them. So, if you're committed to reiser3 or xfs, nothing is stopping you from continuing to deploy systems using them.
3) Many benchmarks show reiser3 as performing better than ext3, and this is mostly true. What isn't shown in those benchmarks is that if you're operating two or more reiser3 file systems in parallel, performance will degrade for both of them due to the use of BKL everywhere. ext3 (and other file systems) will don't degrade in that case. I've also read reports that there is a bit of research going on into making ext3 locking finer grained. I don't have any sources to cite, but any reduction of critical sections without reducing reliability is always a good thing.
People refer to reiser3 as a modern file system, but I'd call it progressive. Reiser3 has served us well for years, but it's showing its age. The basic idea behind reiser3 is still sound, and when extended with integrated integrity checking and better b-tree locking borrowed from years of database research, it would perform extremely well. The problem is that adding the first is a huge disk format change, which means it's no longer reiser3. Adding the second is a hugely invasive change that would throw out a good chunk of the existing code -- again, essentially creating a new file system. It would be like people saying, "I like my ext3 file system, but I don't like the code. Let's start over." Combined with a small development community, it's a recipe for instability and there are more interesting problems out there.
I've posted some more lengthy comments here: http://linux.wordpress.com/2006/09/27/suse-102-di
Rubbish (Score:4, Insightful)
Rubbish. Ext3 has never been able to match Reiser's performance on small files or in other areas, and the notion that ext3 is going to match it is absurd. Even ext4 is not likely to catch up. A lot of ext developers have bizarre ideas about how their filesystem compares to Reiser, XFS or even JFS in a lot of areas. Ext is simply a stable and solid, but badly evolved, filesystem and it is a filesystem that generates an awful lot of disk activity.
I can relate (Score:4, Informative)
Resierfs looked like the clear winner for two good reasons:
1. Reiserfs is faster. Much faster than ext3 in nearly every scenario. Large files and small files.
2. No inode problems. If your users fill your HD with hundreds of thousands of tiny files you're not going to run out of inodes before you run out of disk space. This is something that needs to be anticipated (at the cost of more disk space) at filesystem creation time in ext3.
Reliability for both filesystems was pretty much the same from all accounts.
But in the end I went with ext3 for one and only one reason: Recoverability.
Reiserfs had no, or very few decent, recovery utilities. If a filesystem corruption occurred (and it seemed that the probablity of such corruptions was equal for both filesystems), then data on an ext3 fs stands a much better chance of being recovered than on a reiserfs one.
Of course that was late 2005; that situation may have changed by now.
Re:It's Deja Vu All over Again (Score:5, Insightful)
This was modded flamebait.
People, you might not want to hear it, and you might not agree with stupid knee-jerk reactions, but these reactions will be coming. The name "reiserfs" is tainted, whether that's rational or not.
Regards,
--
*Art
Parent
Re:It's Deja Vu All over Again (Score:4, Insightful)
The other concern is going to be about support, if Hans is found guilty or not, it doesn't really matter. A company such as Novell may consider that the filesystem platform isn't as supported as what it once was and is moving away from it.
From a marketing point of view, Novell won't want to associated with it either. If they show support for him, and he is found guilty, it's a marketing nightmare for Novell.
Parent
Re:It's Deja Vu All over Again (Score:5, Funny)
Are you kidding - this stuff practically writes itself.
Parent
Not just the name... (Score:5, Insightful)
Parent
Re:Just rename it (Score:4, Funny)
Parent
Hard to Believe (Score:5, Informative)
Parent
Re:It's Deja Vu All over Again (Score:5, Funny)
Apparently it is to be called "icefs" in Etch.
Something to do with Hans not being available to QA patches by the Debian kernel team.
Parent
Re:It's Deja Vu All over Again (Score:5, Insightful)
This is why when the story hit I posed legitimate questions regarding the filesystem's future (and got flamed for it, BTW, here and on linuxquestions); a person's career work should be viewed independently of his or her personal misdeeds. Otherwise, we should abandon electricity and incandescent lights (Edison was a bit of a bastard, and his invention of the electric chair "tainted" AC), jets (Heinkel was a nazi), Mercury and Apollo programs should never have happened (Wernher von Braun, the brain behind those programs, was a nazi, willing or otherwise). There are many, many worthwhile inventions proposed, designed, and/or implemented by evil people, and yet we use them on a daily basis, because regardless of the creators' nature, philosophy, or misdeeds, they have produced some worthwhile things that abandoning them because of the heritage would be somewhere between silly and irresponsible.
Parent
Re:Let this be a lesson to you all (Score:5, Funny)
Ok, ReiserFS might be a bit on the slow and unstable side, but I would not actually call it a crime.
Parent
Re:conversion (Score:5, Funny)
Parent