Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?

How To Implement A Database Oriented File System 232

ALundi writes "A really great read from Andrew Orlowski over at The Register on how Benoit Schillings and Dominic Giampaolo created the 64-bit journaled and attribute based Be File System. Schillings and Giampaolo discuss a variety of design and implementation issues, including data integrity and file system performance. " Interesting in the context of MSFTs plans to implement a DB filesystem in future versions of MS Windows.
This discussion has been archived. No new comments can be posted.

How To Implement A Database Oriented File System

Comments Filter:
  • by morgajel ( 568462 ) <slashreader AT morgajel DOT com> on Saturday March 30, 2002 @02:54PM (#3254726) Homepage
    I used Be on and off for about 6 months. Once you get the hang of it(the filesystem), you see the true power- especially with the address book.
    If you want some real insights into the OS as a whole, check out the BEOS Bible []. not so good if you want an in depth discussion, but for non-kernel hackers it's a fun read and very informative.
    Read the section explaining how their address book works. it's really cool.
  • by redhatbox ( 569534 ) <> on Saturday March 30, 2002 @02:54PM (#3254727)

    If I remember correctly, Be originally used a "true" database backend as the filesystem, but ran into performance issues compared to the R5 fs implementation. I can't help but wonder how many of these issues were largely due to the speed of the technology used at the time.

    I suppose it depends on your application, but I know a lot of web-based platforms already use true db backends (Oracle, PostgreSQL) to handle all data storage, representing a filesystem as a hierarchial set of rows in numerous tables. I've written several applications this way, and am currently working on a content management platform which also uses this model. Need to make a change to the filesystem structure (adding attributes, changing the security model)? Just modify the DB structure and you're done, especially with databases like PostgreSQL where you can use the database engine itself for a *lot* of functions (via triggers, stored procedures, security settings, etc).

    As more and more functionality is brought into web-based application environments, I can see the importance of "old style" filesystems starting to fade somewhat for a lot of apps. Yes, they'll still be necessary (the database itself has to reside somewhere, obvisouly), but not in the same way they used to be. Just a few thoughts :).

    • If you were using Oracle, you could take advantage of iFS [] to provide a file system integrated with your database.

      It's a true (NFS, SMB) mountable filesystem, with all the flexibility that provides, but also the replication and other Enterprise features you want from a database.

      • I checked over the specs for this over at the Oracle site... looks impressive! Now, if only you didn't have to use Oracle to take advantage of this sort of thing; PostgreSQL is standing in first place as my database platform of choice for most tasks.

        Is there any kind of equivalent work being done in the OSS community, perhaps based on PGSQL? The platform I'm working on for web-based content management is actually headed toward being usable (and will be licensed under the GPL as soon as that happens), but doesn't use the "real" system FS for storage at all. BLOBs are used for file storage instead, and performance seems to be fairly good :). Of course, now that I've said that the Gods will strike me down...

        I suppose the best way to go is to structure the system such that it can be used for intranet-style doc/file management, in addition to being able to dish up web-based content (whether for internal or external access/use). We'll see how it shapes up, I guess :).

        Hardware has come a long way since the days of BeOS 4 (I believe that's the last version that used a "true" database backend for the filesystem). With the added power we have these days, it seems the old performance issues might be largely eliminated. Now, I'm not trying to portray myself as an advocate of "throwing hardware at a software problem"; to me, it's really more a case of "we can now do things we couldn't do before." This is somewhat akin to being able to do better 3D modelling on PCs due to new capabilities in 3D hardware acceleration.

  • A lot of people think journaling is a really difficult, complicated thing. But that was actually the easiest part by far. BFS journaling is maybe a thousand, maybe twelve hundred lines of code it was really not difficult. And people make it into this monstrous complicated thing. But again, we do things like change the disk buffer cache so the 64bit features that were needed to do journaling were supported.

    In SGI's XFS, their journaling is bigger than all of BFS!

    If its so easy, why don't all file systems implement such goodness? Personally, I'd love to see this everywhere.

    The Gardener

    • Because it changes things. Programs that wipe data no longer work, because the data no longer goes to the same place. Basically, it makes it harder to hack on the system and understand it.
      While it's great for performance, it changes how people understand filesystems. This isn't necessarily a bad thing, it's just a change.
      • Journalling great for performance? I thought performance was what you lose in the tradeoff?
        • It depends.

          SGI's XFS uses and agressive advanced cache system and pre-allocates sectors. What this means is that small files, like temp and working files, never actually make it to the physical disk. They are cached and used from the cache. If they are short lived, then no disk I/O occurs at all.

          Pre-allocating sectors means the system can study where the best location for the file is before it actually gets written. This not only minimized fragmentation, but it also speeds up writes.

  • by entrylevel ( 559061 ) <> on Saturday March 30, 2002 @02:57PM (#3254750)

    ...while Dominic, we are delighted to learn, has subsequently joined Apple as a file system engineer. (He started last week)...

    Does this mean that Apple is finally going to put some kind of reasonably modern filesystem under OS X?

    Have they finally seen the true genius behind their own iTunes interface?

    Have they finally realized that they will shortly be THE ONLY operating system that still relies on file extensions as the primary way of identifying files?

    I truly hope that this snippet is as wonderful as it sounds, as it may finally restore my faith in Apple, as well as cure me of my unhealthy Debian and XFS addiction.
    • Does this mean that Apple is finally going to put some kind of reasonably modern filesystem under OS X?
      You obviously haven't read much about the HFS+ Volume Format []. It's very interesting and even extensible, the problem is none of the advanced features of it are being used by Apple.

      The real beauty of the BeFS wasn't the journaling (that's been done before) it was the ability to let the user define their own meta-data. That's the sort of thing Apple really needs to add to Mac OS X (that and eliminating extensions and Type/Creator in favour of MIME types).

      • You obviously haven't read much about the HFS+ Volume Format []. It's very interesting and even extensible, the problem is none of the advanced features of it are being used by Apple.

        Actually I am quite aware of the capabilities of HFS+, but at the same time I am very disappointed in the use those capabilities are getting. It would be a small feat (for a filesystem engineer, not me) to add completely extensible metadata support and basic journalling to HFS+. Instead, Apple (at the moment) seems to be foresaking every good aspect of the filesystem except for long file names. Perhaps because they realize the performance/reliability they would gain by moving all the metadata out of the filesystem into a RDBMS.

        In reponse to "what's wrong with file extensions?": nothing, unless they are used as the sole method of identifiying files. While this isn't quite the case in OS X, the guidelines seem to state (once again, at the moment) that it will be.

        There is nothing wrong with file extensions, just the same as there is nothing wrong with the creator/type system. However, when you use one and not the other, you are losing capabilities.

        File extensions allow any system, regardless of file system, to tell what a file is supposed to be. Creators and types allow the user's computer to know what application the user wants the file to open in. Magic-number checks tell the computer definitively what the file contains.

        I'm not biased towards or against any of these systems. I want the best consumer-oriented UNIX in the world to have them all!
    • Have they finally realized that they will shortly be THE ONLY operating system that still relies on file extensions as the primary way of identifying files?

      They still have Metadata, to an extent.
    • Have they finally realized that they will shortly be THE ONLY operating system that still relies on file extensions as the primary way of identifying files?

      I think MacOS uses data they store in the resource fork of a file to determine what type of file it is and what application created it. This allows different files of the same type to be opened by different applications, and doesn't require *any* file extension.
    • What are you talking about? HFS+ can store metadata, and OS X can use Creator/Type information. Apple has added the ability to use file extensions to make interoperation with you Windows/Unix knuckle-draggers easier, but the metadata capabilities are still there.
    • Have they finally realized that they will shortly be THE ONLY operating system that still relies on file extensions as the primary way of identifying files?

      You gotta love how every Unix/Linux/Windows user now talks about how file extensions are so bad. A few years ago these same people used to say how crap Apple's idea of meta data was, and that file extensions were better (cross platform, etc, etc). The arguments that I've had with Unix people at different places I've worked. It seems to take the rest of the world at least 10 years to catch on to ideas...
  • Dominic's Book (Score:5, Informative)

    by jimm ( 5532 ) <> on Saturday March 30, 2002 @02:59PM (#3254764) Homepage
    Dominic's book, Practical File System Design with the Be File System, is wonderful. I'd never delved into the innards of a file system before. Reading his book was enjoyable and interesting. I learned quite a lot.
  • MS is have trouble getting major clients to switch to Active Directory. I think there is little chance a DB base FS will be accepted in any reasonable time frame.
    • MS is have (sic) trouble getting major clients to switch to Active Directory.

      Damn right they are! For good reason too: it's cranky and fussy and likes to corrupt itself. When the school I went to threw the Microsoft Official Courseware labs out the window because they were impossible to implement, I got my first taste of why AD as it stands is pretty much useless.

      If Microsoft had stayed standards-compliant with open standards like LDAP and Kerberos 5 and so forth AD would be much less of a nightmare than it is now. But no, typical MS, they had to "embrace and extend" it. As a consequence, they have shot themselves in the foot.

      This is the reason why most MS shops hold desperately on to their NT4 PDCs even though 2K has NT4 beat nine ways to Sunday. 2K cannot do the old-fashioned SAM-based domain even if you cajole it, beat it about the head and shoulders, or ask it nicely. And for most shops, that kind of domain is all they need.

      Of course if they went with Samba they could decommission their old fugly NT4 PDC, heh heh...

      • All of that means that all corporation will eventually be forced to migrate to AD whether they like it or not. How corporations pay to get their options taken away from them and make themselves bitches for MS never ceases to amaze me. The CIOs of america are awfully fond of saying "thank you sir may I have another!".
      • Active Directory *is* LDAP, and they use standard Kerberos for authentication. You can use an OpenLDAP client to query every bit of information a Windows 2000 domain controller holds, and use MIT Kerberos to authenticate against it.

        Where is that standard-incompliant?
  • Database-like (Score:5, Insightful)

    by fm6 ( 162816 ) on Saturday March 30, 2002 @03:03PM (#3254787) Homepage Journal
    It's important to note that they ended up with something "database-like" rather than a true relational DBMS. That distinction is often overlooked (not least by MySQL enthusiasts!) and is pretty important. The thought of a workstation file system that has all the performance and maintenance issues of a "real" DBMS strikes me as pretty scary.

    XFS [] is also "database-like". But BFS seems to be rather more ambititous an effort -- and very intriguing.

    This is one of several BeOS features that the Open Source community should reall consider stealing. But let's consider these features individually, with one eye on whether they're likely to achieve acceptance outside the ranks of BeOS enthusiasts. Let's not waste time on wholesale BeOS clones and compatibility layers. Those are exercises in denial. BeOS was a nice piece of work, but it's as dead as CP/M. Deal with it.

    • This is one of several BeOS features that the Open Source community should reall consider stealing. But let's consider these features individually, with one eye on whether they're likely to achieve acceptance outside the ranks of BeOS enthusiasts. Let's not waste time on wholesale BeOS clones and compatibility layers. Those are exercises in denial. BeOS was a nice piece of work, but it's as dead as CP/M. Deal with it.

      Yes, BeOS is dead. Long live OpenBeOS. []
      • <sarcasm>Great! You bought a domain name! You started a web site! This guarantees that you will have all the volunteer development talent you need! And once you have an OS, everybody will just abandon Windows and Linux and hurry to install OpenBeOS -- just as they did with BeOS! But are you prepared for all the competition from OS/2, DR DOS, NextStep, and AmigaDOS? </sarcasm>

        Look, there's a lot of good stuff in BeOS, and a lot of us would like to see it preserved. But it's the technology that's worth preserving. The platform is just a way of delivering the technology. If you insist on totally re-inventing the platform, either as an OS or as a compatibility layer, you force all your potential users to start over completely. And they just won't do it. You want proof? OS/2. DR DOS. NextStep. AmigaDOS. All of these were impressive products. I personally would prefer any of them (or BeOS for that matter) to Windows. I might even prefer some of them to Linux. But like most users, I have to concentrate on platforms that have a real user base.

        • You are either trolling, or are very ignorant of OpenBeOS.

          It was started many months ago, already has 150+ people working on it, already has a kernel in progress (based on NewOS kernel, written by an ex-Be engineer), and is aiming for BINARY COMPATABILITY with BeOS for the first release. Most of your argument doesn't make any sense if you know all the facts, so instead of just criticizing the project, why not read the website FAQ's.

          P.S. Whoever said we are creating OpenBeOS to try and take over the OS world? We like BeOS, we want to re-create it in OpenBeOS, and then once we have something on par with R5 BeOS, we can go forward with new ideas (which are already being hashed out in a separate sourceforge project called The Glass Elevator).

          • Whoever said we are creating OpenBeOS to try and take over the OS world? We like BeOS, we want to re-create it in OpenBeOS, and then ... we can go forward with new ideas ...
            That's very nice. But what is your goal? A viable software product isn't just a code base. It's a significant community of users applying the product to real-world problems. BeOS simply failed to create that community, and I haven't heard any compelling arguments that OpenBeOS will do any better.

            You're right, I'm ignorant as to how many people are working on OpenBeOS. There are thousands of open source projects doing interesting things. I can't even follow all the ones that are closely-related to my own work. I'll simply take your word for it that a lot of people are working on this particular one.

            But if that's true, I'm all the more discouraged. What is the goal to all this work? Just to avoid moving your desktops to Linux or OS X? Seems rather a lot of work devoted to a very limited goal. The sad fact is, you're actually working against having Be technology becoming more widely accepted.

            Maybe I'm just whining because I'd really like to play with some of this technology. But as long as this technology is only available in the form of a totally new platform, I don't have any room for it. My job already requires me to spend time with more platforms than I care to think about: various desktop Linux distros, Windows, Symbian, and .NET. My personal interests include PalmOS and Java. There's simply no time for me to integrate a whole new platform into my job or other work. (If there were, it would probably be one of the embedded or game-console Linuxes.) And I'm more flexible than most users, most of whom are totally resistent to trying new platforms.

            But if a Linux implementation of the Be file system or the Be file type model were to appear, I'd jump on it with both feet. I might even lobby for my companies development products to support them. But I guess you don't want that to happen.

        • It's worth noting that not all of the OpenBeOS team think in terms of "Give me Be or give me an abacus!" In particular, the file system team [] makes a point of describing their work as "a kernel add-on". Well, I guess that's architecture they inherited from BeOS. But one member of the team, Will Dyson, is also working on a Linux driver. Which I intend to try out at the first opportunity. The possibilities of integration with KDE are very intriguing.
  • Recently to take attention away from Be, microsoft has been bragging about there database'd operating system. According to a microsoft rep, "our operating systems have allways been database'd, everything in windows is based on data".
  • Aren't filesystems by itself hierarchical databases? What's all the fuss about?
    • by rycamor ( 194164 )
      And heirarchical databases were replaced in the 70s by the relational database model because it was impossible to effectively deal with data if you are restricted to a simple heirarchical view.

      Here's an analogy: Heirarchical databases are to relational databases as the GOTO statement is to object encapsulation.

      This might explain why some of us get so frustrated at being forced to continually "navigate" up and down our "folder" heirarchies with the "tree widget". That's just a graphical metaphor for 60s computer technology.
  • wrong way round (Score:2, Interesting)

    by DrSkwid ( 118965 )
    for the majority of databases the data should be moved to the filesystem no the database.

    Simple joins, and most of them are can be replicated with links if necessary. Almost all the databases I've seen would lose little from moving out of the DB and into the filesystem.

    It doesn't scale to complicated joins and huge datastores with complex triggering but for most stuff it simply isn't used.

    Too many developers have the mindset of placing tree based data into RDBMS which adds complexity.

    • Re:wrong way round (Score:5, Interesting)

      by JordanH ( 75307 ) on Saturday March 30, 2002 @03:21PM (#3254924) Homepage Journal
      • for the majority of databases the data should be moved to the filesystem no the database.

        Simple joins, and most of them are can be replicated with links if necessary. Almost all the databases I've seen would lose little from moving out of the DB and into the filesystem.

      But then, you lose all the advantages of database journaling (not just integrity, the ability to rollback to a previous state, if necessary), consistency (you have to make sure the files can't be accessed by other operations before the commits are done) and replication.

      If I were building an application today with Oracle, I would be very tempted to use iFS for these reasons.

      • But then, you lose all the advantages of database journaling (not just integrity, the ability to rollback to a previous state, if necessary)

        Tis true, again for the majority of dbases this simply isn't a problem anyway. Week after week I am presented with projects that have "we store the data in a MySQL database" when there really is no operational need for such a system.

        Plan9 and Hurd with synthetic file systems and filters also present an extension to the simple disk based file system. I'm not for a minute suggesting thrwoing all that indexing and query optimisation away where it's needed.

        Rememebr the ask slashdot a while ago about "how can i keep track of my photographs". People were seriously suggesting Oracle and MySQL. All of his data would fit on a floppy and proably into the memory of my microwave oven or car stereo!

    • Surely thou jesteth! open() has to be one of the most expensive unix system calls! What you're proposing would move slower than cold molasses.

  • Possibilities (Score:3, Insightful)

    by Justen ( 517232 ) on Saturday March 30, 2002 @03:15PM (#3254881) Homepage Journal
    The possibilities with the Be file system were pretty much infinite. There were little things in the OS that would show you this, the address book and email "client" being some great examples. It was incredibly powerful. The only problem was when transporting files from one operating system to the other. And, even in this, Be did an admirable job.

    My fear, and I think the reality is that Microsoft will not be so kind.

  • About time eh? Perhaps next we'll have languages that just deal with persistent data directly, without any 'database interface' stuff to code.

    A DB as a file system could still be a mixed blessing though - anyone tried to store code in a database, or other files with lots of different versions that may have different structures? Generally, DBs are weak when it comes to namespaces (like directories), versions (except in some special cases features for time series) and 'schema evolution' (changing the data structure).

    Nothing impossible though - I think Oracle were getting there with the Internet File System, so would be nice to have a PostgreSQL FS in Linux to start playing with!

    Problem is that file systems don't care about data structure and so don't care about changes to that structure, so you can have half a dozen different versions of your address book floating around
  • by WetCat ( 558132 ) on Saturday March 30, 2002 @03:23PM (#3254946)
    Object-oriented graph system
    - need much less overhead than database system
    - can hold multiple ways to access one object, easing semantic link
    - can be secure by allowing only the workflow links.
    It should be organised as a set of object that hold data within them,
    with links between them.
  • Leveraging a DB FS (Score:5, Interesting)

    by bihoy ( 100694 ) on Saturday March 30, 2002 @03:24PM (#3254949)
    One of the areas that I am interested in is using a DB FS with various "helper" apps. These helper apps would provide a way of managing your data. They could be integrated into another app directly or as a plugin or they could be standalone apps by themselves.

    One could provide a helper app that allows you to look at the stored files in ways other than your typical file listing. To do this would require various metadata attributes to be associated with the data. Helper apps would be provided for all of the standard applications that read and write to data to the file.

    For example, one app could set attributes to categorize data. One could then search for data on your system, and potentially others, much in the same way as you would search for stuff on Yahoo.

    Say you are in a rush to finish your taxes but you need to put together an itemized list of business expenses. You have information on them stored in various places including text files, e-mail, spreadsheets, etc. You could use find and grep to go poking around looking for them or you could use a helper app that does a quick search of attributes and presents you with a list of candidates and ca even call up other apps or services to look at the data. Once you have identified the data another attribute can be set that your tax software uses to record it and pull it into your tax forms.
    • hey. im working on something like this. the page for the project is here []. once i get the documentation done you should be able to write the interface in qt or whatever. i currently have a web front end. check out the test drive.
    • Say you are in a rush to finish your taxes...

      C'mon, don't pretend to ignore the #1 application of a filesystem with built-in metadata support: PR0N. :)

      [X] Asian
      [] Blonde
      [X] Blow
      [] No Audio
      [X] Quality: Great!
      [] Quality: Average
      [] Quality: Crap! I should delete it!
      a filename can only encode so much...


  • mySQL filesystem (Score:2, Interesting)

    by yomahz ( 35486 )
    Here's [] a past discussion and here's [] how it's done.

    Has anyone actually tried it?
  • Databases in File systems..wonder why we never need a File System Administrator ? Or is this just a new way to create jobs ???Give me a decent file system which can survive crashes ...

    btw, wont databases make the filesystem slow ?

    • Re:Stop this! (Score:1, Informative)

      by Anonymous Coward
      A transactional database is very good at surviving system crashes because it is completely journalled. Accessing the data out of the database would probably be no faster, but searching on it would be considerably faster.
  • A few weeks ago, a DB filesystem was looked upon by slashdoters as the dumbest idea ever because MS wanted to implement it.

    All of a sudden it's interesting, perhaps all OS's should have a DB file system installed by default.
  • . "Users will be able to start dealing with data as they would with any type of SQL query, so it will no longer matter where the files are located. Organizing data becomes far simpler."
    How does it make it any simpler ? May be faster .. Anyway locate ang glimpse always work for me ..That makes more sense than putting a DB in the FS
  • I'll wait for SP2 (Score:2, Insightful)

    by Bowfinger ( 559430 )
    Interesting in the context of MSFTs plans to implement a DB filesystem in future versions of MS Windows.

    Even more than normal Microsoft bashing, this sounds like a huge challenge for MS to get right. I can't imagine that they'll manage to retrofit a DB filesystem and make it peform adequately on the first try. It would be tough enough even if they started with a clean slate and a small, independent team of top talent. Unfortunately, although they do have some first-class developers, they also have tremendous legacy baggage and a group-think culture.

    I'll wait for at least Service Pack 2 before I put any real data on an MS DB filesystem.

  • Unix lags (Score:2, Insightful)

    by yggdrazil ( 261592 )
    Unix filesystems lags terribly. They don't store the MIME-type. They don't store the preferred app to open a file in. They don't store metadata like the artist and song name if the song happens to be an mp3. They don't have the ability to add gps postion metadata on my digital camara photos. Searches are horribly slow. All unices use different directory conventions. You can't uninstall apps by just moving its icon to the trash (except on osx). App preferences are stored in all sorts of different ways (except on osx).

    Linux is a nice remake of a legacy os, but is hardly the future.

    The open source community needs a good object storage to base a more futureproof os on. Badly. (And a way better UI than XWindows can give us.)
    • I agree that the filesystems on most Unix boxes are pretty dated. But that's a feature of a particular Unix component, not a basic feature of the OS. Unix (and Linux, which is basically a kind of Unix) is simply not tied to any particular file system.

      If you're a fan of the Be file system [] (and I may be turning into one myself), note that you don't have to use BeOS to use the Be file system [].

  • OpenBeOS BFS (Score:3, Informative)

    by Anonymous Coward on Saturday March 30, 2002 @03:40PM (#3255061)
    Bruno G. Albuquerque and Axel Dorfler are working to make a 100% compatiable clone of the file system that shipped with the later versions of BeOS as part of the OpenBeOS [] project.

    They have a fantastic amount of the work done [] already such as :
    o Read-only BFS
    o Kernel Interface
    o Full Attribute Support
    o Indexing
    o Symlink Traversal
    o Queries, full UTF-8 Support, and support for non-indexed attributes.

    I believe they are also fixing some problems that were in the original FS.

    I am sure they would be glad for some more file system engineers. Come to think of it, the rest of this open source project is going really well too, but as always it needs more programmers....

    • Re:OpenBeOS BFS (Score:2, Informative)

      by BGA ( 28170 )
      Just a small correction. Right now we have write support working and files and directories can be created. Growing a file only works for the direct range (that means we didn't even touch the indirect and double indirect range yet). Also, we don't have any journaling in place yet, but we do have a transaction class that *IS* being used (only that it just writes the blocks directly to disk right now without any journaling).
  • A great book on the BeOS file system, its design and the issues involved, is Practical File System Design with the Be File System [] by Dominic Giampaolo. This was the first book I read about file systems and is a great discussion on the types of decisions the BeOS folks made to address the specific users and OS functionality they were looking for. It's not too heavy and is really a casual read in comparison to most other Computer Science books.
  • Is there any work being done to build a database file system for Linux and other os's? I know I have a hell of a time navigating through my mp3s and it would sure be nice to be able to assign attributes to them(rock+favorite+2002) so you could just search on 'favorites and rock' for instance and have a playlist. A database filesystem should also simplify the task of managing files since you could sort them by adding attributes.
  • by Oggust ( 526634 ) <> on Saturday March 30, 2002 @04:28PM (#3255416) Homepage
    But there are some very real problems:

    • Portability: Have you ever tried to move data from DB system to another? Not fun! There need to be some standards! There is such a thing as SQL-92, but nobody uses it yet. Yet? Right...
    • Portablility again: Isn't it nice today how you can get a tar file from just anybode, and it will just untar on your system, even if you haven't got the same OS and filesystem as the guy you got it from? Well see the previous point.
    • Performance, but that's possibly not that big a deal: A database can do a lot of work "server side". What do you think will happen to the system load on a big multi-user system when some moron submits a huge SQL query with all kinds of weird joins and stuff?

      Not much user, lots of system and iowait, that's what. We run into a whole new realm of needing accounting for these kinds of things.


    • That's actually a good point. If you think about it, filesystem use is one of the few things that is more-or-less the same from platform to platform. Sure, directory demarcations change, and line endings, but that's about it. Using the standard C library on just about any system gives you easy compatibility.

      Imagine if you had to really work to abstract your filesystem layer in all of your crossplatform software the way that you have to abstract your windowing system and most other stuff.
  • my two cents (Score:3, Interesting)

    by joshuaos ( 243047 ) <ouroboros.freedoment@com> on Saturday March 30, 2002 @04:37PM (#3255482) Journal
    Benoit: Either you make the data shared, or you make the application components shared. And sharing the data was easier. (Dominic agrees).

    If the data is shared, and you have libraries that are shared, then why not ask the data to display itself (object.display(x);) and have it call to a standard library (system library?) which queries a system properties database object as to what application to display it? Don't actually store the display code in the objects, but have the objects query the system as to what the user has specified to display that type of data with.

    Dominic: That's what I mean. Some people are very anal about organizing things in rigid hierarchies and others are 'I know what I want to find'.

    I think there is a place for hierarchies, but not as the base organizational method of the filesystem. I would like to see a hierarchy of attributes, or keys, or whatever you want to call them. When you save an object (off the internet, or out of your head), a title is only one possible attribute you want to give it. When I save a pr0n jpg, it doesn't need a damn title, I need to mark what it's a damn picture of (amateur AND cumshots AND redhead)! Perhaps start with people, places, things. Or later in the hierarchy, sound -> music -> various bands as well as various artists as well as various sound effects as well as dates and live or studio, all keyed (so to speak) and queryable. But the hierarchy is for browsing. Just for browsing, because browsing is important (when you want to look at cumshots, you want to look at cumshots, but when you query for cumshots, watersports and lesbians, well that's bloody well what you should get), and micro$oft's nice little explorer looks about right. Although instead of a stupid directory tree, we have a tree of object properties and types, and any object can be in any number of places in that tree, depending on it's attributes (categories?).

    I know of course that I haven't really said anything new in this post, and I know that performace needs to be taken into account. This is, however, the way things are moving, and all we really need is a really good, really fast, solid state storage medium. When permanent storage is as fast as or faster than RAM is today, the database filesystem will finally become a reality, until then, we'll sure be gearing up.

    Cheers, Joshua

  • Making the file system a database didn't start with Be or Microsoft, it goes back several decades to IBM.

    There are brute force ways of doing it: you build some kind of ad hoc database system and dump it into kernel space. You may be able to engineer such a system reasonably well, but to me, it is in bad taste: indexing is such a complex and application dependent area that nobody can guess ahead of time what kind of indexing people will want a few years from now. The Be file system looks like it's too complicated to interoperate well, and too simplistic to be of much use for anything rather than fairly primitive indexing operations.

    A better way of doing this is to figure out a protocol for notification and updates between a traditional file system and user-space database indexing services. Yes, that's harder, but that's what software engineers get paid to figure out. And, as far as I'm concerned, if you can't figure out how to do it right, it's better not to do it at all rather than doing something half-baked.

  • Newforge has this commentary [] by James Treleaven about the possible implications to Open Source if Microsoft implements a database driven filesystem.
    • That's a tempest in a teapot. MS Office files still need to be serializable somehow, and that format cannot possibly be worse than it is right now. If anything, it may get better with Microsoft's push for XML.

      Functionally, database-based file systems are an old hat. If they were the magic bullet Microsoft and BeOS think they are, they would have caught on long ago. To me, it looks more like a bunch of college hackers getting mightily excited about a whizbang feature of little real value. (Database based file systems have worked well in some niche markets--IBM is selling some systems with such file systems.)

      Something needs to be done about indexing and search, but putting a database into the kernel is not the right thing.

  • by Anonymous Coward on Saturday March 30, 2002 @05:22PM (#3255801)
    I should know - I actually interviewed for a position in the group developing this. (I signed the standard pre-interview NDA, so I'm posting anonymously, but I'll still try to stay general so I don't get into too much trouble...)

    As the ExtremeTech article pointed out, they are not even considering putting the full-blown SQL Server into Windows. SQL Server is too resource-intensive (it really wants to use all of the available CPU, memory and disk space), too much overhead, and most importantly to MS, too profitable (sales of SQL Server / BackOffice make up about 10-15% of MS's revenue.) There's no reason to bundle it if people are willing to pay a ton for it separately!

    As the article says, they're thinking (nothing decided yet) about including MSDE, which is exactly the same as SQL Server 2000, except it is tuned for 5 concurrent users (and hard-limited to 10), the database size is limited to 2GB per database (the same as the Jet DB, aka Access), and it doesn't have the nice GUI admin tools bundled.

    Also, the OFS (Object File System) discussed previously probably won't get added either. There's a good reason why it was talked about way back in the Cairo (pre-Win95) days but never implemented - it's really, really hard to do, and it's hard to even convince anybody of its value. (Just look at Be.) Active Directory was originally supposed to be an object store, but I don't think anybody uses it for that (if anybody even uses it at all.)

    What probably will be included is an improved version of Indexing Service, which is currently included in Windows 2000 and XP. For those of who are fortunate enough to be unfamiliar with Indexing Service (formerly Index Server), it's an NT service (think "daemon") that periodically scans the file system for new / updated files, and then adds whatever metadata it can extract into a database of sorts, which is then used to speed up searches in the built-in Search dialog on the Start menu.
    There are a couple of problems with the current implementation:
    1. It's extremely buggy. For example, it's supposed to run only when the system is idle, but it usually chooses to run right when I'm doing some big IO-intensive process. It periodically also consumes all available memory and CPU cycles, so I usually leave it turned off.
    2. It doesn't index very much metadata. For most files, it only indexes the standard file system attributes (size, date, extension, etc.) For MS Office documents, it can index the Summary Information property stream, which usually includes things like Author, Title, etc. If you have IIS running on a box, it can also index the META tags in HTML documents. Other than that, not much. It does have some extensibility , by means of DLLs that implement the IFilter COM interface, but no 3rd-party vendor seems to have done much with that, perhaps because of #1.

    So, in summary, MS's plans for the DB-in-the-filesystem look a lot more like Reiser4 [] than like BeFS or SQL Server.
    • The indexing system in W2K is just about useless. I have it turned on and when I ask it to find a file given it's name it still takes over a minute to return answers. On linux or freebsd type in "locate filaname" and you get your answer instantly. I don't know what it does but it sure as hell can't find files fast. Like almost everything MS makes "it makes a lot of promises but delivers almost nothing"
      • by horza ( 87255 )
        I agree that searching for files under Windows is so slow that I find rapid clicking around in folders yields a better average success time. I have to admit that I cannot countenance using a *nix box without locate. Especially when piped through to grep. Want to find the executable for application foobar? locate foobar | grep bin. Result comes back after a few milliseconds. Locate actually builds a database, usually rebuilt 4am each morning by cron, which is the reason it is so fast. If you've just installed a package and need to find where it has just put something then type "locate -u" to update db (though my sys admin friend tells me off for doing this as there is apparently a better way, my way works fine though) and then use locate as usual.

    • they are not even considering putting the full-blown SQL Server into Windows. SQL Server is too resource-intensive.
      I find this very hard to believe. True, SQLServer eats up everything you feed it, but stays away from the stuff you don't give it: if you forbid it to use a certain CPU, it won't use it; if you tune it to only use XX% of the memory , it will only use that amount of memory (and databases use memory mainly for cache).

      they're thinking (nothing decided yet) about including MSDE, which is exactly the same as SQL Server 2000, except it is tuned for 5 concurrent users (and hard-limited to 10)
      MSDE is limited to 5 concurrent transactions, not users. So every 6th transaction (i.e. a T-SQL command) will get queued and has to wait for an empty slot. This will degrade performance but you can hook up 30 or more users on a single MSDE powered database without noticing it.

      Your point about 'including' MSDE and not SQLServer is weird: both share the same codebase and both eat up everything you give them. MSDE too eats up all the memory it can get, even for 5 transactions. Also: a DB Filesystem eats resources, plain and simple, because you have more data to store plus you're not just serving files but VIEWS on bits of data. So caching has to be intensive. Yes this costs memory and perhaps a lot of diskspace.

      As for Index Server: Index Server is ment to index textdocuments and documents of a certain format (office docs etc). It's a hog, indeed, but it also doesn't need much maintenance. It does it's own housekeeping, indexes files by itself. Asking a query by the windows search is a bit slow perhaps, but I've never had any performance problem when using indexserver as the searchengine on static websites.
  • The PICK OS was an even earlier example of a "database filesystem". I worked with it back in 1981, but it dates back earlier than that.

    PICK has an SQL-like language to create reports and search and select databases. The data structures are kept in a "dictonary" file for each file. Records are variable length and seperated by upper acsii characters. (254 for fields, 253 for values (sub-fields), and 252 for sub-values.)

    PICK is pretty much considered a "legacy system" by most people any more. (If they have heard of it at all.) It had some features that were far ahead of its time. Unfortuantly, Dick Pick, the creator of the OS, was unwilling to improve on it. So PICK remained in the dumb terminal world, while everyone else moved on. The vendors of PICK made improvements over the years. It was the only thing that kept it even vaguely current.

    Microsoft is claiming that you will not need datastructures for their new system. DB filesystems are very dependant on embeded filestructures. It has to be there. I have no idea how they will be able to take a structure as complex as Word or MPEG-4 and make it "transparent" and portable. More likely you will be held hostage to their OS. It will be portable to "upgrades" of the OS and no more. (At least until they make that data format no longer supported...)
  • by octogen ( 540500 ) <> on Saturday March 30, 2002 @05:43PM (#3255955)
    It's really not Microsoft's innovation.

    IBM's AS/400 (a midrange computer system targeted for commercial use/accounting/warehouse/etc...) is based on an object-oriented database filesystem which is implemented at the firmware level (SLIC) rather than at the OS-level - and this system has been around for about 20 years and IIRC it always had quite good performance.


    A few words about its architecture, if you're interested...

    The operating system (OS/400) itself runs on top of this object-oriented low-level "OS" by calling its APIs - as a result, most parts of OS/400 are platform-independent. If you'd manage to get the SLIC running on another hardware platform, you could probably install a nearly unmodified version of OS/400, and it would do its work.

    Actually, I'd call the SLIC code the 'real' operating system kernel rather than OS/400, because OS/400 itself would not work without an apropriate SLIC layer.

    Everything on the system is an object, so you'll always have to use the object's methods to perform some operation.
    For some applications that may be an advantage, because security is enforced on each object at the firmware level. For other applications it might also be a disadvantage, because you'll always have to use a limited set of APIs for modifying data. That blocks many methods commonly used for writing highly optimized code.

    -end arch----

    One of the benefits of having a database-filesystem is probably the fact that you do not need to run a database product on top of the OS.
    Every object on the system can be backed up and restored in a very simple way. Logical files (multiple logical views of one physical file) can help to keep data management simple and consistent.

    On the other hand, you will have to update the entire OS (including the kernel) when you need to install a new release of the database - which means, that you'll have to reboot the machine.

    And - last but not least - the more code you have in the OS kernel, the higher is the probability of having dangerous bugs somewhere in the kernel.
    It should not be necessary to mention, that bugs in the OS kernel may compromise all system security.

    There are certainly many advantages and disadvantages regarding the database-filesystem issue, so I think it all depends on what you want to do with your computer.


    kind regards from Austria,

    PS: i hope my english isn't too poor..
    And - by the way - even Microsoft uses AS/400 boxes for running its business, so what do you think, where did they get their inspiration from...?

    • Mod this up!

      This guy has it right. I have been working AS/400's for YEARS! The OS is built around the DB2 database. And IBM beat MS to the idea a long time ago. But, MS likes to make everyone think they actually "Innovate" things. When in fact they don't! They buy up companies, steal technoligies etc... The only thing MS has perfected is the marketing engine. And yes it is true MS uses AS/400's. In fact MS wanted me to come interview to work on their team that did the MS--> As/400 integration work. (I told them to blow me, I wouldn't work for a company I dispised!).

    • As I see it, working with OS/400 is as if you're inside oracle or SQL Server: A table in these DBMS'es is a file in OS/400. Nothing special about that.

      What's the point in MS' DB filesystem is that you're not INSIDE a database, but ON TOP of the database. So you handle files and bits of data as you are doing today but they're not physical files but views (or query results as you wish) on the data in the database. That's something different than a filesystem with metadata connected to the 'files' like in the article, or an OS that is just a DMBS like OS/400.

      You could simulate the behaviour today if you store different pieces of a file in a database (SQL Server, Oracle) as blobs with metadata stored in related fields/tables. Query on the database, f.e. by selecting blobs based on a certain where clause on the metadata, and build some component that will reconstruct a file from the blobs. _THAT_'s what's all about.
    • [Damn, I hate IE's text edit bugs.]

      Dominic, we are delighted to learn, has subsequently joined Apple as a file system engineer.

      This is the best news that I've had in a while on the Mac OSX technical front.

      Allright, Dominic! Get in there and kick some ass!
  • Oracle [] has a database based filesystem Internet File System/IFS [] that is pretty sweet.

    Supports protocols HTTP, WebDAV, SMB, FTP, IMAP, and SMTP gives the ability to store, manage, and search documents, presentations, multimedia, Web pages, and XML files.

    Also do Check-in/check-out, version control, and event notification features. Unlike most of the examples i have seen developers can extend these base features to build custom content management applications.
    • IFS is pretty intriguing. But it's also pretty complex. Oracle products tend to have a huge cost-of-ownership, and IFS looks to be particularly bad. No only do you need to set up an Oracle relational server, you have to put a lot of work into defining mapings between the database and your other entities. Probably not worth the work unless your project is very big.

      In anycase, IFS is not a workstation filesystem, like BFS. Strictly for servers.

  • ... and not just dev nodes in the filesystem.

    EROS [] claims to checkpoint the memory of the system - its running state - rather than have programs explicitly serialize state to save and so on. By doing this they can make a machine shut down and reboot in a matter of seconds into a consistent state with your programs apparently still running. This is different from hibernation - you can take the machine down at any time by pulling the plug and this will work.

    Its a different axis of filesystem from the 'store more metadata' stuff in this thread, but IMHO its a more interesting way to go. EROS's storage is transactional - to the extent that what is on disk always presents a consistent picture of what was in memory; user-level transactions are dealt with in programs. Its unclear from the register article whether the Be filesystem had this level of integrity.


Logic is the chastity belt of the mind!