Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 1 | |
| 2 | Ext4 Filesystem |
| 3 | =============== |
| 4 | |
Diego Calleja | 22359f5 | 2008-10-17 09:15:14 -0400 | [diff] [blame] | 5 | Ext4 is an an advanced level of the ext3 filesystem which incorporates |
| 6 | scalability and reliability enhancements for supporting large filesystems |
| 7 | (64 bit) in keeping with increasing disk capacities and state-of-the-art |
| 8 | feature requirements. |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 9 | |
Diego Calleja | 22359f5 | 2008-10-17 09:15:14 -0400 | [diff] [blame] | 10 | Mailing list: linux-ext4@vger.kernel.org |
| 11 | Web site: http://ext4.wiki.kernel.org |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 12 | |
| 13 | |
| 14 | 1. Quick usage instructions: |
| 15 | =========================== |
| 16 | |
Diego Calleja | 22359f5 | 2008-10-17 09:15:14 -0400 | [diff] [blame] | 17 | Note: More extensive information for getting started with ext4 can be |
| 18 | found at the ext4 wiki site at the URL: |
| 19 | http://ext4.wiki.kernel.org/index.php/Ext4_Howto |
| 20 | |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 21 | - Compile and install the latest version of e2fsprogs (as of this |
Diego Calleja | 22359f5 | 2008-10-17 09:15:14 -0400 | [diff] [blame] | 22 | writing version 1.41.3) from: |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 23 | |
| 24 | http://sourceforge.net/project/showfiles.php?group_id=2406 |
| 25 | |
| 26 | or |
| 27 | |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 28 | ftp://ftp.kernel.org/pub/linux/kernel/people/tytso/e2fsprogs/ |
| 29 | |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 30 | or grab the latest git repository from: |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 31 | |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 32 | git://git.kernel.org/pub/scm/fs/ext2/e2fsprogs.git |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 33 | |
Theodore Ts'o | 4537398 | 2008-07-27 19:59:21 -0400 | [diff] [blame] | 34 | - Note that it is highly important to install the mke2fs.conf file |
| 35 | that comes with the e2fsprogs 1.41.x sources in /etc/mke2fs.conf. If |
| 36 | you have edited the /etc/mke2fs.conf file installed on your system, |
| 37 | you will need to merge your changes with the version from e2fsprogs |
| 38 | 1.41.x. |
| 39 | |
Theodore Ts'o | 03010a3 | 2008-10-10 20:02:48 -0400 | [diff] [blame] | 40 | - Create a new filesystem using the ext4 filesystem type: |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 41 | |
Theodore Ts'o | 03010a3 | 2008-10-10 20:02:48 -0400 | [diff] [blame] | 42 | # mke2fs -t ext4 /dev/hda1 |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 43 | |
Diego Calleja | 22359f5 | 2008-10-17 09:15:14 -0400 | [diff] [blame] | 44 | Or to configure an existing ext3 filesystem to support extents: |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 45 | |
Diego Calleja | 22359f5 | 2008-10-17 09:15:14 -0400 | [diff] [blame] | 46 | # tune2fs -O extents /dev/hda1 |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 47 | |
| 48 | If the filesystem was created with 128 byte inodes, it can be |
| 49 | converted to use 256 byte for greater efficiency via: |
| 50 | |
| 51 | # tune2fs -I 256 /dev/hda1 |
| 52 | |
Theodore Ts'o | 03010a3 | 2008-10-10 20:02:48 -0400 | [diff] [blame] | 53 | (Note: we currently do not have tools to convert an ext4 |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 54 | filesystem back to ext3; so please do not do try this on production |
| 55 | filesystems.) |
| 56 | |
| 57 | - Mounting: |
| 58 | |
Theodore Ts'o | 03010a3 | 2008-10-10 20:02:48 -0400 | [diff] [blame] | 59 | # mount -t ext4 /dev/hda1 /wherever |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 60 | |
| 61 | - When comparing performance with other filesystems, remember that |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 62 | ext3/4 by default offers higher data integrity guarantees than most. |
| 63 | So when comparing with a metadata-only journalling filesystem, such |
| 64 | as ext3, use `mount -o data=writeback'. And you might as well use |
| 65 | `mount -o nobh' too along with it. Making the journal larger than |
| 66 | the mke2fs default often helps performance with metadata-intensive |
| 67 | workloads. |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 68 | |
| 69 | 2. Features |
| 70 | =========== |
| 71 | |
| 72 | 2.1 Currently available |
| 73 | |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 74 | * ability to use filesystems > 16TB (e2fsprogs support not available yet) |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 75 | * extent format reduces metadata overhead (RAM, IO for access, transactions) |
| 76 | * extent format more robust in face of on-disk corruption due to magics, |
| 77 | * internal redunancy in tree |
Mingming Cao | 49f1487 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 78 | * improved file allocation (multi-block alloc) |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 79 | * fix 32000 subdirectory limit |
| 80 | * nsec timestamps for mtime, atime, ctime, create time |
| 81 | * inode version field on disk (NFSv4, Lustre) |
| 82 | * reduced e2fsck time via uninit_bg feature |
| 83 | * journal checksumming for robustness, performance |
| 84 | * persistent file preallocation (e.g for streaming media, databases) |
| 85 | * ability to pack bitmaps and inode tables into larger virtual groups via the |
| 86 | flex_bg feature |
| 87 | * large file support |
| 88 | * Inode allocation using large virtual block groups via flex_bg |
Mingming Cao | 49f1487 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 89 | * delayed allocation |
| 90 | * large block (up to pagesize) support |
| 91 | * efficent new ordered mode in JBD2 and ext4(avoid using buffer head to force |
| 92 | the ordering) |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 93 | |
| 94 | 2.2 Candidate features for future inclusion |
| 95 | |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 96 | * Online defrag (patches available but not well tested) |
| 97 | * reduced mke2fs time via lazy itable initialization in conjuction with |
| 98 | the uninit_bg feature (capability to do this is available in e2fsprogs |
| 99 | but a kernel thread to do lazy zeroing of unused inode table blocks |
| 100 | after filesystem is first mounted is required for safety) |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 101 | |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 102 | There are several others under discussion, whether they all make it in is |
| 103 | partly a function of how much time everyone has to work on them. Features like |
| 104 | metadata checksumming have been discussed and planned for a bit but no patches |
| 105 | exist yet so I'm not sure they're in the near-term roadmap. |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 106 | |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 107 | The big performance win will come with mballoc, delalloc and flex_bg |
| 108 | grouping of bitmaps and inode tables. Some test results available here: |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 109 | |
Diego Calleja | 22359f5 | 2008-10-17 09:15:14 -0400 | [diff] [blame] | 110 | - http://www.bullopensource.org/ext4/20080818-ffsb/ffsb-write-2.6.27-rc1.html |
| 111 | - http://www.bullopensource.org/ext4/20080818-ffsb/ffsb-readwrite-2.6.27-rc1.html |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 112 | |
| 113 | 3. Options |
| 114 | ========== |
| 115 | |
| 116 | When mounting an ext4 filesystem, the following option are accepted: |
| 117 | (*) == default |
| 118 | |
Alex Tomas | c9de560 | 2008-01-29 00:19:52 -0500 | [diff] [blame] | 119 | extents (*) ext4 will use extents to address file data. The |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 120 | file system will no longer be mountable by ext3. |
| 121 | |
Alex Tomas | c9de560 | 2008-01-29 00:19:52 -0500 | [diff] [blame] | 122 | noextents ext4 will not use extents for newly created files |
| 123 | |
Girish Shilamkar | 818d276 | 2008-01-28 23:58:27 -0500 | [diff] [blame] | 124 | journal_checksum Enable checksumming of the journal transactions. |
| 125 | This will allow the recovery code in e2fsck and the |
| 126 | kernel to detect corruption in the kernel. It is a |
| 127 | compatible change and will be ignored by older kernels. |
| 128 | |
| 129 | journal_async_commit Commit block can be written to disk without waiting |
| 130 | for descriptor blocks. If enabled older kernels cannot |
| 131 | mount the device. This will enable 'journal_checksum' |
| 132 | internally. |
| 133 | |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 134 | journal=update Update the ext4 file system's journal to the current |
| 135 | format. |
| 136 | |
| 137 | journal=inum When a journal already exists, this option is ignored. |
| 138 | Otherwise, it specifies the number of the inode which |
| 139 | will represent the ext4 file system's journal file. |
| 140 | |
| 141 | journal_dev=devnum When the external journal device's major/minor numbers |
| 142 | have changed, this option allows the user to specify |
| 143 | the new journal location. The journal device is |
| 144 | identified through its new major/minor numbers encoded |
| 145 | in devnum. |
| 146 | |
| 147 | noload Don't load the journal on mounting. |
| 148 | |
| 149 | data=journal All data are committed into the journal prior to being |
| 150 | written into the main file system. |
| 151 | |
| 152 | data=ordered (*) All data are forced directly out to the main file |
| 153 | system prior to its metadata being committed to the |
| 154 | journal. |
| 155 | |
| 156 | data=writeback Data ordering is not preserved, data may be written |
| 157 | into the main file system after its metadata has been |
| 158 | committed to the journal. |
| 159 | |
| 160 | commit=nrsec (*) Ext4 can be told to sync all its data and metadata |
| 161 | every 'nrsec' seconds. The default value is 5 seconds. |
| 162 | This means that if you lose your power, you will lose |
| 163 | as much as the latest 5 seconds of work (your |
| 164 | filesystem will not be damaged though, thanks to the |
| 165 | journaling). This default value (or any low value) |
| 166 | will hurt performance, but it's good for data-safety. |
| 167 | Setting it to 0 will have the same effect as leaving |
| 168 | it at the default (5 seconds). |
| 169 | Setting it to very large values will improve |
| 170 | performance. |
| 171 | |
Eric Sandeen | 571640c | 2008-05-26 12:29:46 -0400 | [diff] [blame] | 172 | barrier=<0|1(*)> This enables/disables the use of write barriers in |
| 173 | the jbd code. barrier=0 disables, barrier=1 enables. |
| 174 | This also requires an IO stack which can support |
| 175 | barriers, and if jbd gets an error on a barrier |
| 176 | write, it will disable again with a warning. |
| 177 | Write barriers enforce proper on-disk ordering |
| 178 | of journal commits, making volatile disk write caches |
| 179 | safe to use, at some performance penalty. If |
| 180 | your disks are battery-backed in one way or another, |
| 181 | disabling barriers may safely improve performance. |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 182 | |
Theodore Ts'o | 240799c | 2008-10-09 23:53:47 -0400 | [diff] [blame] | 183 | inode_readahead=n This tuning parameter controls the maximum |
| 184 | number of inode table blocks that ext4's inode |
| 185 | table readahead algorithm will pre-read into |
| 186 | the buffer cache. The default value is 32 blocks. |
| 187 | |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 188 | orlov (*) This enables the new Orlov block allocator. It is |
| 189 | enabled by default. |
| 190 | |
| 191 | oldalloc This disables the Orlov block allocator and enables |
| 192 | the old block allocator. Orlov should have better |
| 193 | performance - we'd like to get some feedback if it's |
| 194 | the contrary for you. |
| 195 | |
| 196 | user_xattr Enables Extended User Attributes. Additionally, you |
| 197 | need to have extended attribute support enabled in the |
| 198 | kernel configuration (CONFIG_EXT4_FS_XATTR). See the |
| 199 | attr(5) manual page and http://acl.bestbits.at/ to |
| 200 | learn more about extended attributes. |
| 201 | |
| 202 | nouser_xattr Disables Extended User Attributes. |
| 203 | |
| 204 | acl Enables POSIX Access Control Lists support. |
| 205 | Additionally, you need to have ACL support enabled in |
| 206 | the kernel configuration (CONFIG_EXT4_FS_POSIX_ACL). |
| 207 | See the acl(5) manual page and http://acl.bestbits.at/ |
| 208 | for more information. |
| 209 | |
| 210 | noacl This option disables POSIX Access Control List |
| 211 | support. |
| 212 | |
| 213 | reservation |
| 214 | |
| 215 | noreservation |
| 216 | |
| 217 | bsddf (*) Make 'df' act like BSD. |
| 218 | minixdf Make 'df' act like Minix. |
| 219 | |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 220 | debug Extra debugging information is sent to syslog. |
| 221 | |
| 222 | errors=remount-ro(*) Remount the filesystem read-only on an error. |
| 223 | errors=continue Keep going on a filesystem error. |
| 224 | errors=panic Panic and halt the machine if an error occurs. |
| 225 | |
Hidehiro Kawai | 5bf5683 | 2008-10-10 22:12:43 -0400 | [diff] [blame] | 226 | data_err=ignore(*) Just print an error message if an error occurs |
| 227 | in a file data buffer in ordered mode. |
| 228 | data_err=abort Abort the journal if an error occurs in a file |
| 229 | data buffer in ordered mode. |
| 230 | |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 231 | grpid Give objects the same group ID as their creator. |
| 232 | bsdgroups |
| 233 | |
| 234 | nogrpid (*) New objects have the group ID of their creator. |
| 235 | sysvgroups |
| 236 | |
| 237 | resgid=n The group ID which may use the reserved blocks. |
| 238 | |
| 239 | resuid=n The user ID which may use the reserved blocks. |
| 240 | |
| 241 | sb=n Use alternate superblock at this location. |
| 242 | |
| 243 | quota |
| 244 | noquota |
| 245 | grpquota |
| 246 | usrquota |
| 247 | |
| 248 | bh (*) ext4 associates buffer heads to data pages to |
| 249 | nobh (a) cache disk block mapping information |
| 250 | (b) link pages into transaction to provide |
| 251 | ordering guarantees. |
| 252 | "bh" option forces use of buffer heads. |
| 253 | "nobh" option tries to avoid associating buffer |
| 254 | heads (supported only for "writeback" mode). |
| 255 | |
Alex Tomas | c9de560 | 2008-01-29 00:19:52 -0500 | [diff] [blame] | 256 | stripe=n Number of filesystem blocks that mballoc will try |
| 257 | to use for allocation size and alignment. For RAID5/6 |
| 258 | systems this should be the number of data |
| 259 | disks * RAID chunk size in file system blocks. |
Mingming Cao | 49f1487 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 260 | delalloc (*) Deferring block allocation until write-out time. |
| 261 | nodelalloc Disable delayed allocation. Blocks are allocation |
| 262 | when data is copied from user to page cache. |
Theodore Ts'o | 240799c | 2008-10-09 23:53:47 -0400 | [diff] [blame] | 263 | |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 264 | Data Mode |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 265 | ========= |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 266 | There are 3 different data modes: |
| 267 | |
| 268 | * writeback mode |
| 269 | In data=writeback mode, ext4 does not journal data at all. This mode provides |
| 270 | a similar level of journaling as that of XFS, JFS, and ReiserFS in its default |
| 271 | mode - metadata journaling. A crash+recovery can cause incorrect data to |
| 272 | appear in files which were written shortly before the crash. This mode will |
| 273 | typically provide the best ext4 performance. |
| 274 | |
| 275 | * ordered mode |
| 276 | In data=ordered mode, ext4 only officially journals metadata, but it logically |
Mingming Cao | 49f1487 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 277 | groups metadata information related to data changes with the data blocks into a |
| 278 | single unit called a transaction. When it's time to write the new metadata |
| 279 | out to disk, the associated data blocks are written first. In general, |
| 280 | this mode performs slightly slower than writeback but significantly faster than journal mode. |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 281 | |
| 282 | * journal mode |
| 283 | data=journal mode provides full data and metadata journaling. All new data is |
| 284 | written to the journal first, and then to its final location. |
| 285 | In the event of a crash, the journal can be replayed, bringing both data and |
| 286 | metadata into a consistent state. This mode is the slowest except when data |
| 287 | needs to be read from and written to disk at the same time where it |
Mingming Cao | 49f1487 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 288 | outperforms all others modes. Curently ext4 does not have delayed |
| 289 | allocation support if this data journalling mode is selected. |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 290 | |
| 291 | References |
| 292 | ========== |
| 293 | |
| 294 | kernel source: <file:fs/ext4/> |
| 295 | <file:fs/jbd2/> |
| 296 | |
| 297 | programs: http://e2fsprogs.sourceforge.net/ |
Dave Kleikamp | fc513a3 | 2006-10-11 01:21:25 -0700 | [diff] [blame] | 298 | |
| 299 | useful links: http://fedoraproject.org/wiki/ext3-devel |
| 300 | http://www.bullopensource.org/ext4/ |
Jose R. Santos | 93e3270 | 2008-07-11 19:27:31 -0400 | [diff] [blame] | 301 | http://ext4.wiki.kernel.org/index.php/Main_Page |
| 302 | http://fedoraproject.org/wiki/Features/Ext4 |