DOCSLIB.ORG

Embedded File Systems (Part 4) COLUMNS

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text

Load more

54 CIRCUIT CELLAR • APRIL 2014 #285 EMBEDDED IN THIN SLICES Embedded File Systems (Part 4) COLUMNS Specialized Linux Tools Several types of file systems are available for Linux. This article discusses compressed read- only file systems (CRAMFS), RAM file systems, network file systems (NFS), and the Samba networking protocol. By Bob Japenga (US) o help embedded systems designers using the built-in zlib compression to enable Tachieve their design goals, Linux offers random access of all of the files. This makes a variety of file systems that require only a CRAMFS relatively fast. The file metadata moderate amount of effort to implement. This (e.g., information about when the file was article introduces several file systems and a created, read and write privileges, etc.) is networking protocol and describes how and not compressed but uses a more compact when to use each one. notation than is present in most file systems. WHAT IS A CRAMFS? WHEN TO USE A CRAMFS While pursuing my Master’s degree in The primary reason my company has used 1973, a professor made a statement I will a CRAMFS is to cut down on the flash memory never forget: “Memory prices are plunging used by the file system. The first embedded so fast you will never have to worry about Linux system we worked on had 16 MB of RAM how much memory your designs use from and 32 MB of flash. There was a systems- now on.” Obviously, this professor did not level requirement to provide a means for have a crystal ball and missed something the system to recover should the primary that has been a design challenge in virtually partition become corrupt or fail to boot every embedded system I have designed. Our in any way (see Part 3 of this article series systems demand more and more memory (or “Designing Robust Flash Memory Systems,” file space) and a compressed read-only file Circuit Cellar 283, 2014, for more detail). We system (CRAMFS) can be a useful solution in met this requirement by creating a backup some instances. partition that used a CRAMFS. A CRAMFS is an open-source file system The backup partition’s only function was to available for Linux. I am not sure where the enable the box to recover from the corrupted CRAMFS gets its name. Perhaps it gets its primary partition. How we accomplished that name because a CRAMFS is one way to cram is a discussion for another day, but we had your file system into a smaller footprint. very little flash memory to hold the Linux The files are compressed one page at a time kernel, the file system, the boot-loader, and circuitcellar.com 55 the application. We were able to have the two mount command, you specify the ramfs size. file systems identical in file content, which However, it can grow in size to exceed that made it easy to maintain. Using a CRAMFS amount of RAM. Thus ramfs will enable you enabled us to cut our backup file system to use your entire RAM and not provide you space requirements in half. any with warning that it is doing it. tmpfs A second feature of a CRAMFS is its read- does not enable you to write more than the only nature. Given that it is read-only, it space allocated at mount time. An error is does not require wear leveling. This keeps returned when you try to use more than you the overhead for using CRAMFS files very have allocated. Another difference is that low. Due to the typical data retention of flash tmpfs uses swap space and can swap seldom memory, this also means that for products used files out to a flash drive. ramfs does that will be deployed for more than 10 years, not use swapping. This difference is of little you will need to rewrite the CRAMFS partition value to us since we disable swapping in our COLUMNS every three to five years. embedded systems. The fact that the file system is read-only can make it difficult to deploy in all situations. WHEN TO USE RAM FILE SYSTEMS For example, once our CRAMFS began using In 1987, the Department of Defense began encryption keys that were uniquely calculated requiring the use of Ada in new embedded for each system one time at start-up (this systems. Shortly thereafter I bought an Ada took several minutes on a 200-mHz ARM9 compiler for my IBM PC at work. Much to my processor) the CRAMFS became a problem. surprise, the compiler came with a board We could not store the calculated keys on to plug into the PC. The board added a RAM the CRAMFS and the start-up delay was file system to my PC. To improve the speed unacceptable. This forced us to come up with of compiling from glacial to turtle-like, the an alternate backup file system. compiler designers used a RAM file system to store all intermediate results. HOW TO MAKE A CRAMFS Speed is one of the primary reasons to use As with the other file systems I have a RAM file system. Disk writes are lightning discussed, implementing a CRAMFS with fast when you have a RAM disk. We have used Linux is simple and relatively painless. Linux a RAM file system when we are recording a provides a tool (mkfs) you can use to take any burst of high-speed data. In the background, group of directories and files and create a we write the data out to flash. CRAMFS image. Actually, mkfs is a wrapper A second reason to use a RAM file system for a variety of file system building tools. is that it reduces the wear and tear on the mkfs.cramfs is the tool used to create the flash file system, which has a limited write CRAMFS image. Once written, the image is life. We make it a rule that all temporary files placed in its own partition. should be kept on the RAM disk. We also use it for temporary variables that are needed WHAT ARE RAM FILE SYSTEMS? across threads/processes. Linux provides two types of RAM file systems: ramfs and tmpfs. Both are full- HOW TO MAKE A RAM FILE SYSTEM featured file systems that reside in RAM and Since tmpfs is built into Linux, you can are thus very fast and volatile (i.e., the data simply mount a RAM drive with a command is not retained across power outages and such as: system restarts). When the systems are created with the mount – o size=16G –t tmpfs my_ Hard drive CRC PLC PLC Embedded Embedded Embedded Embedded PLC Embedded Embedded Embedded Embedded controller controller controller controller controller controller controller controller FIGURE 1 This layout shows the memory regions Embedded Embedded Embedded Embedded required to bring up one of my controller controller controller controller company’s embedded Linux designs with no redundancy. 56 CIRCUIT CELLAR • APRIL 2014 #285 tmpfs /mnt/tmpfs The major area in which we use NFS today is during software development. Even though where the maximum size is 16 GB, the type flash file systems are fast and new versions of of file system is tmpfs, the name is my_ your code can be seamlessly written to flash, tmpfs, and the access point (device name) is it can be time consuming. For example, you /mnt/tmpfs. The size can also be specified can use a flash memory stick over USB to in terms of percentage of your total physical update the flash file system on several of our RAM. tmpfs will only use the RAM that it designs. This is simple but can take anywhere needs up to the maximum specified. If you from several seconds to minutes. don’t specify a maximum, it defaults to 50% With an NFS, all of your development tools of your total physical RAM. can be on a PC and you never have to transfer ramfs is also built into Linux and is created the target code to the target system. You use with a command such as: all of your PC tools to change the file on your ABOUT THE AUTHOR PC, and when the embedded device boots up mount –o size=20m –t ramfs my_ Bob Japenga has been or the application is restarted, those changed COLUMNS ramfs /mnt/ramfs files will be used on the device. designing embedded systems since 1973. In where the type is ramfs, the name is my_ HOW TO MAKE AN NFS 1988, along with his ramfs and the access point (device name) is An NFS is built into Linux and thus is pretty best friend, he start- /mnt/ramfs. trivial to set up. The embedded system can ed MicroTools, which mount an external drive using a combination specializes in creating WHAT ARE NETWORK FILE of the Mount command and an entry in the file a variety of real-time SYSTEMS? system table file (usually in /etc/fstab). The embedded systems. In the early 1990s I started working with a external drive requires an NFS driver that is With a combined em- company that developed embedded controllers easily available for Windows and is built into bedded systems experi- for machine control. These controllers had a all Linux distributions. ence base of more than user interface that consisted of a PC located 200 years, they love to on the factory floor. The company called WHAT IS SAMBA? tackle impossible prob- this the production line console (PLC). The Although we don’t like to admit it, many lems together.
Recommended publications
  • Study of File System Evolution

    Study of File System Evolution

    Study of File System Evolution Swaminathan Sundararaman, Sriram Subramanian Department of Computer Science University of Wisconsin {swami, srirams} @cs.wisc.edu Abstract File systems have traditionally been a major area of file systems are typically developed and maintained by research and development. This is evident from the several programmer across the globe. At any point in existence of over 50 file systems of varying popularity time, for a file system, there are three to six active in the current version of the Linux kernel. They developers, ten to fifteen patch contributors but a single represent a complex subsystem of the kernel, with each maintainer. These people communicate through file system employing different strategies for tackling individual file system mailing lists [14, 16, 18] various issues. Although there are many file systems in submitting proposals for new features, enhancements, Linux, there has been no prior work (to the best of our reporting bugs, submitting and reviewing patches for knowledge) on understanding how file systems evolve. known bugs. The problems with the open source We believe that such information would be useful to the development approach is that all communication is file system community allowing developers to learn buried in the mailing list archives and aren’t easily from previous experiences. accessible to others. As a result when new file systems are developed they do not leverage past experience and This paper looks at six file systems (Ext2, Ext3, Ext4, could end up re-inventing the wheel. To make things JFS, ReiserFS, and XFS) from a historical perspective worse, people could typically end up doing the same (between kernel versions 1.0 to 2.6) to get an insight on mistakes as done in other file systems.
  • ECE 598 – Advanced Operating Systems Lecture 19

    ECE 598 – Advanced Operating Systems Lecture 19

    ECE 598 { Advanced Operating Systems Lecture 19 Vince Weaver http://web.eece.maine.edu/~vweaver [email protected] 7 April 2016 Announcements • Homework #7 was due • Homework #8 will be posted 1 Why use FAT over ext2? • FAT simpler, easy to code • FAT supported on all major OSes • ext2 faster, more robust filename and permissions 2 btrfs • B-tree fs (similar to a binary tree, but with pages full of leaves) • overwrite filesystem (overwite on modify) vs CoW • Copy on write. When write to a file, old data not overwritten. Since old data not over-written, crash recovery better Eventually old data garbage collected • Data in extents 3 • Copy-on-write • Forest of trees: { sub-volumes { extent-allocation { checksum tree { chunk device { reloc • On-line defragmentation • On-line volume growth 4 • Built-in RAID • Transparent compression • Snapshots • Checksums on data and meta-data • De-duplication • Cloning { can make an exact snapshot of file, copy-on- write different than link, different inodles but same blocks 5 Embedded • Designed to be small, simple, read-only? • romfs { 32 byte header (magic, size, checksum,name) { Repeating files (pointer to next [0 if none]), info, size, checksum, file name, file data • cramfs 6 ZFS Advanced OS from Sun/Oracle. Similar in idea to btrfs indirect still, not extent based? 7 ReFS Resilient FS, Microsoft's answer to brtfs and zfs 8 Networked File Systems • Allow a centralized file server to export a filesystem to multiple clients. • Provide file level access, not just raw blocks (NBD) • Clustered filesystems also exist, where multiple servers work in conjunction.
  • CIS 191 Linux Lab Exercise

    CIS 191 Linux Lab Exercise

    CIS 191 Linux Lab Exercise Lab 9: Backup and Restore Fall 2008 Lab 9: Backup and Restore The purpose of this lab is to explore the different types of backups and methods of restoring them. We will look at three utilities often used for backups and learn the advantages and disadvantages of each. Supplies • VMWare Server 1.05 or higher • Benji VM Preconfiguration • Labs 6 and Labs 7 completed on a pristine Benji VM [root@benji /]# fdisk -l Disk /dev/sda: 5368 MB, 5368709120 bytes 255 heads, 63 sectors/track, 652 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/sda1 * 1 382 3068383+ 83 Linux /dev/sda2 383 447 522112+ 82 Linux swap / Solaris /dev/sda3 448 511 514080 83 Linux /dev/sda4 512 652 1132582+ 5 Extended /dev/sda5 512 549 305203+ 83 Linux /dev/sda6 550 556 56196 83 Linux /dev/sda7 557 581 200781 83 Linux [root@benji /]# [root@benji /]# mount /dev/sda1 on / type ext3 (rw) proc on /proc type proc (rw) sysfs on /sys type sysfs (rw) devpts on /dev/pts type devpts (rw,gid=5,mode=620) tmpfs on /dev/shm type tmpfs (rw) /dev/sda5 on /opt type ext3 (rw) /dev/sda3 on /var type ext3 (rw) none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw) sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw) /dev/sda7 on /home type ext3 (rw,usrquota) [root@benji /]# cat /etc/fstab LABEL=/1 / ext3 defaults 1 1 devpts /dev/pts devpts gid=5,mode=620 0 0 tmpfs /dev/shm tmpfs defaults 0 0 LABEL=/opt /opt ext3 defaults 1 2 proc /proc proc defaults 0 0 sysfs /sys sysfs defaults 0 0 LABEL=/var /var ext3 defaults 1 2 LABEL=SWAP-sda2 swap swap defaults 0 0 LABEL=/home /home ext3 usrquota,defaults 1 2 [root@benji /]# Forum If you get stuck on one of the steps below don’t beat your head against the wall.
  • Hardware-Driven Evolution in Storage Software by Zev Weiss A

    Hardware-Driven Evolution in Storage Software by Zev Weiss A

    Hardware-Driven Evolution in Storage Software by Zev Weiss A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Computer Sciences) at the UNIVERSITY OF WISCONSIN–MADISON 2018 Date of final oral examination: June 8, 2018 ii The dissertation is approved by the following members of the Final Oral Committee: Andrea C. Arpaci-Dusseau, Professor, Computer Sciences Remzi H. Arpaci-Dusseau, Professor, Computer Sciences Michael M. Swift, Professor, Computer Sciences Karthikeyan Sankaralingam, Professor, Computer Sciences Johannes Wallmann, Associate Professor, Mead Witter School of Music i © Copyright by Zev Weiss 2018 All Rights Reserved ii To my parents, for their endless support, and my cousin Charlie, one of the kindest people I’ve ever known. iii Acknowledgments I have taken what might be politely called a “scenic route” of sorts through grad school. While Ph.D. students more focused on a rapid graduation turnaround time might find this regrettable, I am glad to have done so, in part because it has afforded me the opportunities to meet and work with so many excellent people along the way. I owe debts of gratitude to a large cast of characters: To my advisors, Andrea and Remzi Arpaci-Dusseau. It is one of the most common pieces of wisdom imparted on incoming grad students that one’s relationship with one’s advisor (or advisors) is perhaps the single most important factor in whether these years of your life will be pleasant or unpleasant, and I feel exceptionally fortunate to have ended up iv with the advisors that I’ve had.
  • Connecting the Storage System to the Solaris Host

    Connecting the Storage System to the Solaris Host

    Configuration Guide for Solaris™ Host Attachment Hitachi Virtual Storage Platform Hitachi Universal Storage Platform V/VM FASTFIND LINKS Document Organization Product Version Getting Help Contents MK-96RD632-05 Copyright © 2010 Hitachi, Ltd., all rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or stored in a database or retrieval system for any purpose without the express written permission of Hitachi, Ltd. (hereinafter referred to as “Hitachi”) and Hitachi Data Systems Corporation (hereinafter referred to as “Hitachi Data Systems”). Hitachi Data Systems reserves the right to make changes to this document at any time without notice and assumes no responsibility for its use. This document contains the most current information available at the time of publication. When new and/or revised information becomes available, this entire document will be updated and distributed to all registered users. All of the features described in this document may not be currently available. Refer to the most recent product announcement or contact your local Hitachi Data Systems sales office for information about feature and product availability. Notice: Hitachi Data Systems products and services can be ordered only under the terms and conditions of the applicable Hitachi Data Systems agreement(s). The use of Hitachi Data Systems products is governed by the terms of your agreement(s) with Hitachi Data Systems. Hitachi is a registered trademark of Hitachi, Ltd. in the United States and other countries. Hitachi Data Systems is a registered trademark and service mark of Hitachi, Ltd.
  • Enterprise Filesystems

    Enterprise Filesystems

    Enterprise Filesystems Eric Sandeen Principal Software Engineer, Red Hat Feb 21, 2013 1 ERIC SANDEEN What We'll Cover ● Local “Enterprise-ready” Linux filesystems ● Ext3 ● Ext4 ● XFS ● BTRFS ● Use cases, features, pros & cons of each ● Recent & future work ● Features ● Scalability ● Benchmarks 2 ERIC SANDEEN Local Filesystems in RHEL6 ● We ship what customers need and can rely on ● We ship what we test and support ● Major on-disk local filesystems ● Ext3, Ext4, XFS, BTRFS* ● Others are available for special purposes ● fat, vfat, msdos, udf, cramfs, squashfs... ● We'll cover the “big four” today 3 ERIC SANDEEN The Ext3 filesystem ● Ext3 is was the most common file system in Linux ● Most distributions historically used it as their default ● Applications tuned to its specific behaviors (fsync...) ● Familiar to most system administrators ● Ext3 challenges ● File system repair (fsck) time can be extremely long ● Limited scalability - maximum file system size of 16TB ● Can be significantly slower than other local file systems ● direct/indirect, bitmaps, no delalloc ... 4 ERIC SANDEEN The Ext4 filesystem ● Ext4 has many compelling new features ● Extent based allocation ● Faster fsck time (up to 10x over ext3) ● Delayed allocation, preallocation ● Higher bandwidth ● Should be relatively familiar for existing ext3 users ● Ext4 challenges ● Large device support not polished in its user space tools ● Limits supported maximum file system size to 16TB* ● Has different behavior over system failure 5 ERIC SANDEEN The XFS filesystem ● XFS is very robust
  • Managing File Systems in Oracle® Solaris 11.4

    Managing File Systems in Oracle® Solaris 11.4

    ® Managing File Systems in Oracle Solaris 11.4 Part No: E61016 November 2020 Managing File Systems in Oracle Solaris 11.4 Part No: E61016 Copyright © 2004, 2020, Oracle and/or its affiliates. License Restrictions Warranty/Consequential Damages Disclaimer This software and related documentation are provided under a license agreement containing restrictions on use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license, transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means. Reverse engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is prohibited. Warranty Disclaimer The information contained herein is subject to change without notice and is not warranted to be error-free. If you find any errors, please report them to us in writing. Restricted Rights Notice If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it on behalf of the U.S. Government, then the following notice is applicable: U.S. GOVERNMENT END USERS: Oracle programs (including any operating system, integrated software, any programs embedded, installed or activated on delivered hardware, and modifications of such programs) and Oracle computer documentation or other Oracle data delivered to or accessed by U.S. Government end users are "commercial computer software" or "commercial
  • Unionfs: User- and Community-Oriented Development of a Unification File System

    Unionfs: User- and Community-Oriented Development of a Unification File System

    Unionfs: User- and Community-Oriented Development of a Unification File System David Quigley, Josef Sipek, Charles P. Wright, and Erez Zadok Stony Brook University {dquigley,jsipek,cwright,ezk}@cs.sunysb.edu Abstract If a file exists in multiple branches, the user sees only the copy in the higher-priority branch. Unionfs allows some branches to be read-only, Unionfs is a stackable file system that virtually but as long as the highest-priority branch is merges a set of directories (called branches) read-write, Unionfs uses copy-on-write seman- into a single logical view. Each branch is as- tics to provide an illusion that all branches are signed a priority and may be either read-only writable. This feature allows Live-CD develop- or read-write. When the highest priority branch ers to give their users a writable system based is writable, Unionfs provides copy-on-write se- on read-only media. mantics for read-only branches. These copy- on-write semantics have lead to widespread There are many uses for namespace unifica- use of Unionfs by LiveCD projects including tion. The two most common uses are Live- Knoppix and SLAX. In this paper we describe CDs and diskless/NFS-root clients. On Live- our experiences distributing and maintaining CDs, by definition, the data is stored on a read- an out-of-kernel module since November 2004. only medium. However, it is very convenient As of March 2006 Unionfs has been down- for users to be able to modify the data. Uni- loaded by over 6,700 unique users and is used fying the read-only CD with a writable RAM by over two dozen other projects.
  • We Get Letters Sept/Oct 2018

    We Get Letters Sept/Oct 2018

    SEE TEXT ONLY WeGetletters by Michael W Lucas letters@ freebsdjournal.org tmpfs, or be careful to monitor tmpfs space use. Hey, FJ Letters Dude, Not that you’ll configure your monitoring system Which filesystem should I use? to watch tmpfs, because it’s temporary. And no matter what, one day you’ll forget —FreeBSD Newbie that you used memory space as a filesystem. You’ll stash something vital in that temporary space, then reboot. And get really annoyed Dear FreeBSD Newbie, when that vital data vanishes into the ether. First off, welcome to FreeBSD. The wider com- Some other filesystems aren’t actively terrible. munity is glad to help you. The device filesystem devfs(5) provides device Second, please let me know who told you to nodes. Filesystems that can’t store user data are start off by writing me. I need to properly… the best filesystems. But then some clever sysad- “thank” them. min decides to hack on /etc/devfs.rules to Filesystems? Sure, let’s talk filesystems. change the standard device nodes for their spe- Discussing which filesystem is the worst is like cial application, or /etc/devd.conf to create or debating the merits of two-handed swords as reconfigure device nodes, and the whole system compared to lumberjack-grade chainsaws and goes down the tubes. industrial tulip presses. While every one of them Speaking of clever sysadmins, now and then has perfectly legitimate uses, in the hands of the people decide that they want to optimize disk novice they’re far more likely to maim everyone space or cut down how many copies of a file involved.
  • CIS Ubuntu Linux 18.04 LTS Benchmark

    CIS Ubuntu Linux 18.04 LTS Benchmark

    CIS Ubuntu Linux 18.04 LTS Benchmark v1.0.0 - 08-13-2018 Terms of Use Please see the below link for our current terms of use: https://www.cisecurity.org/cis-securesuite/cis-securesuite-membership-terms-of-use/ 1 | P a g e Table of Contents Terms of Use ........................................................................................................................................................... 1 Overview ............................................................................................................................................................... 12 Intended Audience ........................................................................................................................................ 12 Consensus Guidance ..................................................................................................................................... 13 Typographical Conventions ...................................................................................................................... 14 Scoring Information ..................................................................................................................................... 14 Profile Definitions ......................................................................................................................................... 15 Acknowledgements ...................................................................................................................................... 17 Recommendations ............................................................................................................................................
  • Comparison of Kernel and User Space File Systems

    Comparison of Kernel and User Space File Systems

    Comparison of kernel and user space file systems — Bachelor Thesis — Arbeitsbereich Wissenschaftliches Rechnen Fachbereich Informatik Fakultät für Mathematik, Informatik und Naturwissenschaften Universität Hamburg Vorgelegt von: Kira Isabel Duwe E-Mail-Adresse: [email protected] Matrikelnummer: 6225091 Studiengang: Informatik Erstgutachter: Professor Dr. Thomas Ludwig Zweitgutachter: Professor Dr. Norbert Ritter Betreuer: Michael Kuhn Hamburg, den 28. August 2014 Abstract A file system is part of the operating system and defines an interface between OS and the computer’s storage devices. It is used to control how the computer names, stores and basically organises the files and directories. Due to many different requirements, such as efficient usage of the storage, a grand variety of approaches arose. The most important ones are running in the kernel as this has been the only way for a long time. In 1994, developers came up with an idea which would allow mounting a file system in the user space. The FUSE (Filesystem in Userspace) project was started in 2004 and implemented in the Linux kernel by 2005. This provides the opportunity for a user to write an own file system without editing the kernel code and therefore avoid licence problems. Additionally, FUSE offers a stable library interface. It is originally implemented as a loadable kernel module. Due to its design, all operations have to pass through the kernel multiple times. The additional data transfer and the context switches are causing some overhead which will be analysed in this thesis. So, there will be a basic overview about on how exactly a file system operation takes place and which mount options for a FUSE-based system result in a better performance.
  • State of the Art: Where We Are with the Ext3 Filesystem

    State of the Art: Where We Are with the Ext3 Filesystem

    State of the Art: Where we are with the Ext3 filesystem Mingming Cao, Theodore Y. Ts’o, Badari Pulavarty, Suparna Bhattacharya IBM Linux Technology Center {cmm, theotso, pbadari}@us.ibm.com, [email protected] Andreas Dilger, Alex Tomas, Cluster Filesystem Inc. [email protected], [email protected] Abstract 1 Introduction Although the ext2 filesystem[4] was not the first filesystem used by Linux and while other filesystems have attempted to lay claim to be- ing the native Linux filesystem (for example, The ext2 and ext3 filesystems on Linux R are when Frank Xia attempted to rename xiafs to used by a very large number of users. This linuxfs), nevertheless most would consider the is due to its reputation of dependability, ro- ext2/3 filesystem as most deserving of this dis- bustness, backwards and forwards compatibil- tinction. Why is this? Why have so many sys- ity, rather than that of being the state of the tem administrations and users put their trust in art in filesystem technology. Over the last few the ext2/3 filesystem? years, however, there has been a significant amount of development effort towards making There are many possible explanations, includ- ext3 an outstanding filesystem, while retaining ing the fact that the filesystem has a large and these crucial advantages. In this paper, we dis- diverse developer community. However, in cuss those features that have been accepted in our opinion, robustness (even in the face of the mainline Linux 2.6 kernel, including direc- hardware-induced corruption) and backwards tory indexing, block reservation, and online re- compatibility are among the most important sizing.