linux/Documentation/filesystems/f2fs.txt
<<
>>
Prefs
   1================================================================================
   2WHAT IS Flash-Friendly File System (F2FS)?
   3================================================================================
   4
   5NAND flash memory-based storage devices, such as SSD, eMMC, and SD cards, have
   6been equipped on a variety systems ranging from mobile to server systems. Since
   7they are known to have different characteristics from the conventional rotating
   8disks, a file system, an upper layer to the storage device, should adapt to the
   9changes from the sketch in the design level.
  10
  11F2FS is a file system exploiting NAND flash memory-based storage devices, which
  12is based on Log-structured File System (LFS). The design has been focused on
  13addressing the fundamental issues in LFS, which are snowball effect of wandering
  14tree and high cleaning overhead.
  15
  16Since a NAND flash memory-based storage device shows different characteristic
  17according to its internal geometry or flash memory management scheme, namely FTL,
  18F2FS and its tools support various parameters not only for configuring on-disk
  19layout, but also for selecting allocation and cleaning algorithms.
  20
  21The file system formatting tool, "mkfs.f2fs", is available from the following
  22git tree:
  23>> git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs-tools.git
  24
  25For reporting bugs and sending patches, please use the following mailing list:
  26>> linux-f2fs-devel@lists.sourceforge.net
  27
  28================================================================================
  29BACKGROUND AND DESIGN ISSUES
  30================================================================================
  31
  32Log-structured File System (LFS)
  33--------------------------------
  34"A log-structured file system writes all modifications to disk sequentially in
  35a log-like structure, thereby speeding up  both file writing and crash recovery.
  36The log is the only structure on disk; it contains indexing information so that
  37files can be read back from the log efficiently. In order to maintain large free
  38areas on disk for fast writing, we divide  the log into segments and use a
  39segment cleaner to compress the live information from heavily fragmented
  40segments." from Rosenblum, M. and Ousterhout, J. K., 1992, "The design and
  41implementation of a log-structured file system", ACM Trans. Computer Systems
  4210, 1, 26\xE2\x80\x9352.
  43
  44Wandering Tree Problem
  45----------------------
  46In LFS, when a file data is updated and written to the end of log, its direct
  47pointer block is updated due to the changed location. Then the indirect pointer
  48block is also updated due to the direct pointer block update. In this manner,
  49the upper index structures such as inode, inode map, and checkpoint block are
  50also updated recursively. This problem is called as wandering tree problem [1],
  51and in order to enhance the performance, it should eliminate or relax the update
  52propagation as much as possible.
  53
  54[1] Bityutskiy, A. 2005. JFFS3 design issues. http://www.linux-mtd.infradead.org/
  55
  56Cleaning Overhead
  57-----------------
  58Since LFS is based on out-of-place writes, it produces so many obsolete blocks
  59scattered across the whole storage. In order to serve new empty log space, it
  60needs to reclaim these obsolete blocks seamlessly to users. This job is called
  61as a cleaning process.
  62
  63The process consists of three operations as follows.
  641. A victim segment is selected through referencing segment usage table.
  652. It loads parent index structures of all the data in the victim identified by
  66   segment summary blocks.
  673. It checks the cross-reference between the data and its parent index structure.
  684. It moves valid data selectively.
  69
  70This cleaning job may cause unexpected long delays, so the most important goal
  71is to hide the latencies to users. And also definitely, it should reduce the
  72amount of valid data to be moved, and move them quickly as well.
  73
  74================================================================================
  75KEY FEATURES
  76================================================================================
  77
  78Flash Awareness
  79---------------
  80- Enlarge the random write area for better performance, but provide the high
  81  spatial locality
  82- Align FS data structures to the operational units in FTL as best efforts
  83
  84Wandering Tree Problem
  85----------------------
  86- Use a term, \xE2\x80\x9Cnode\xE2\x80\x9D, that represents inodes as well as various pointer blocks
  87- Introduce Node Address Table (NAT) containing the locations of all the \xE2\x80\x9Cnode\xE2\x80\x9D
  88  blocks; this will cut off the update propagation.
  89
  90Cleaning Overhead
  91-----------------
  92- Support a background cleaning process
  93- Support greedy and cost-benefit algorithms for victim selection policies
  94- Support multi-head logs for static/dynamic hot and cold data separation
  95- Introduce adaptive logging for efficient block allocation
  96
  97================================================================================
  98MOUNT OPTIONS
  99================================================================================
 100
 101background_gc_off      Turn off cleaning operations, namely garbage collection,
 102                       triggered in background when I/O subsystem is idle.
 103disable_roll_forward   Disable the roll-forward recovery routine
 104discard                Issue discard/TRIM commands when a segment is cleaned.
 105no_heap                Disable heap-style segment allocation which finds free
 106                       segments for data from the beginning of main area, while
 107                       for node from the end of main area.
 108nouser_xattr           Disable Extended User Attributes. Note: xattr is enabled
 109                       by default if CONFIG_F2FS_FS_XATTR is selected.
 110noacl                  Disable POSIX Access Control List. Note: acl is enabled
 111                       by default if CONFIG_F2FS_FS_POSIX_ACL is selected.
 112active_logs=%u         Support configuring the number of active logs. In the
 113                       current design, f2fs supports only 2, 4, and 6 logs.
 114                       Default number is 6.
 115disable_ext_identify   Disable the extension list configured by mkfs, so f2fs
 116                       does not aware of cold files such as media files.
 117
 118================================================================================
 119DEBUGFS ENTRIES
 120================================================================================
 121
 122/sys/kernel/debug/f2fs/ contains information about all the partitions mounted as
 123f2fs. Each file shows the whole f2fs information.
 124
 125/sys/kernel/debug/f2fs/status includes:
 126 - major file system information managed by f2fs currently
 127 - average SIT information about whole segments
 128 - current memory footprint consumed by f2fs.
 129
 130================================================================================
 131USAGE
 132================================================================================
 133
 1341. Download userland tools and compile them.
 135
 1362. Skip, if f2fs was compiled statically inside kernel.
 137   Otherwise, insert the f2fs.ko module.
 138 # insmod f2fs.ko
 139
 1403. Create a directory trying to mount
 141 # mkdir /mnt/f2fs
 142
 1434. Format the block device, and then mount as f2fs
 144 # mkfs.f2fs -l label /dev/block_device
 145 # mount -t f2fs /dev/block_device /mnt/f2fs
 146
 147Format options
 148--------------
 149-l [label]   : Give a volume label, up to 512 unicode name.
 150-a [0 or 1]  : Split start location of each area for heap-based allocation.
 151               1 is set by default, which performs this.
 152-o [int]     : Set overprovision ratio in percent over volume size.
 153               5 is set by default.
 154-s [int]     : Set the number of segments per section.
 155               1 is set by default.
 156-z [int]     : Set the number of sections per zone.
 157               1 is set by default.
 158-e [str]     : Set basic extension list. e.g. "mp3,gif,mov"
 159-t [0 or 1]  : Disable discard command or not.
 160               1 is set by default, which conducts discard.
 161
 162================================================================================
 163DESIGN
 164================================================================================
 165
 166On-disk Layout
 167--------------
 168
 169F2FS divides the whole volume into a number of segments, each of which is fixed
 170to 2MB in size. A section is composed of consecutive segments, and a zone
 171consists of a set of sections. By default, section and zone sizes are set to one
 172segment size identically, but users can easily modify the sizes by mkfs.
 173
 174F2FS splits the entire volume into six areas, and all the areas except superblock
 175consists of multiple segments as described below.
 176
 177                                            align with the zone size <-|
 178                 |-> align with the segment size
 179     _________________________________________________________________________
 180    |            |            |   Segment   |    Node     |   Segment  |      |
 181    | Superblock | Checkpoint |    Info.    |   Address   |   Summary  | Main |
 182    |    (SB)    |   (CP)     | Table (SIT) | Table (NAT) | Area (SSA) |      |
 183    |____________|_____2______|______N______|______N______|______N_____|__N___|
 184                                                                       .      .
 185                                                             .                .
 186                                                 .                            .
 187                                    ._________________________________________.
 188                                    |_Segment_|_..._|_Segment_|_..._|_Segment_|
 189                                    .           .
 190                                    ._________._________
 191                                    |_section_|__...__|_
 192                                    .            .
 193                                    .________.
 194                                    |__zone__|
 195
 196- Superblock (SB)
 197 : It is located at the beginning of the partition, and there exist two copies
 198   to avoid file system crash. It contains basic partition information and some
 199   default parameters of f2fs.
 200
 201- Checkpoint (CP)
 202 : It contains file system information, bitmaps for valid NAT/SIT sets, orphan
 203   inode lists, and summary entries of current active segments.
 204
 205- Segment Information Table (SIT)
 206 : It contains segment information such as valid block count and bitmap for the
 207   validity of all the blocks.
 208
 209- Node Address Table (NAT)
 210 : It is composed of a block address table for all the node blocks stored in
 211   Main area.
 212
 213- Segment Summary Area (SSA)
 214 : It contains summary entries which contains the owner information of all the
 215   data and node blocks stored in Main area.
 216
 217- Main Area
 218 : It contains file and directory data including their indices.
 219
 220In order to avoid misalignment between file system and flash-based storage, F2FS
 221aligns the start block address of CP with the segment size. Also, it aligns the
 222start block address of Main area with the zone size by reserving some segments
 223in SSA area.
 224
 225Reference the following survey for additional technical details.
 226https://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey
 227
 228File System Metadata Structure
 229------------------------------
 230
 231F2FS adopts the checkpointing scheme to maintain file system consistency. At
 232mount time, F2FS first tries to find the last valid checkpoint data by scanning
 233CP area. In order to reduce the scanning time, F2FS uses only two copies of CP.
 234One of them always indicates the last valid data, which is called as shadow copy
 235mechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism.
 236
 237For file system consistency, each CP points to which NAT and SIT copies are
 238valid, as shown as below.
 239
 240  +--------+----------+---------+
 241  |   CP   |    SIT   |   NAT   |
 242  +--------+----------+---------+
 243  .         .          .          .
 244  .            .              .              .
 245  .               .                 .                 .
 246  +-------+-------+--------+--------+--------+--------+
 247  | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 |
 248  +-------+-------+--------+--------+--------+--------+
 249     |             ^                          ^
 250     |             |                          |
 251     `----------------------------------------'
 252
 253Index Structure
 254---------------
 255
 256The key data structure to manage the data locations is a "node". Similar to
 257traditional file structures, F2FS has three types of node: inode, direct node,
 258indirect node. F2FS assigns 4KB to an inode block which contains 923 data block
 259indices, two direct node pointers, two indirect node pointers, and one double
 260indirect node pointer as described below. One direct node block contains 1018
 261data blocks, and one indirect node block contains also 1018 node blocks. Thus,
 262one inode block (i.e., a file) covers:
 263
 264  4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB.
 265
 266   Inode block (4KB)
 267     |- data (923)
 268     |- direct node (2)
 269     |          `- data (1018)
 270     |- indirect node (2)
 271     |            `- direct node (1018)
 272     |                       `- data (1018)
 273     `- double indirect node (1)
 274                         `- indirect node (1018)
 275                                      `- direct node (1018)
 276                                                 `- data (1018)
 277
 278Note that, all the node blocks are mapped by NAT which means the location of
 279each node is translated by the NAT table. In the consideration of the wandering
 280tree problem, F2FS is able to cut off the propagation of node updates caused by
 281leaf data writes.
 282
 283Directory Structure
 284-------------------
 285
 286A directory entry occupies 11 bytes, which consists of the following attributes.
 287
 288- hash          hash value of the file name
 289- ino           inode number
 290- len           the length of file name
 291- type          file type such as directory, symlink, etc
 292
 293A dentry block consists of 214 dentry slots and file names. Therein a bitmap is
 294used to represent whether each dentry is valid or not. A dentry block occupies
 2954KB with the following composition.
 296
 297  Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) +
 298                      dentries(11 * 214 bytes) + file name (8 * 214 bytes)
 299
 300                         [Bucket]
 301             +--------------------------------+
 302             |dentry block 1 | dentry block 2 |
 303             +--------------------------------+
 304             .               .
 305       .                             .
 306  .       [Dentry Block Structure: 4KB]       .
 307  +--------+----------+----------+------------+
 308  | bitmap | reserved | dentries | file names |
 309  +--------+----------+----------+------------+
 310  [Dentry Block: 4KB] .   .
 311                 .               .
 312            .                          .
 313            +------+------+-----+------+
 314            | hash | ino  | len | type |
 315            +------+------+-----+------+
 316            [Dentry Structure: 11 bytes]
 317
 318F2FS implements multi-level hash tables for directory structure. Each level has
 319a hash table with dedicated number of hash buckets as shown below. Note that
 320"A(2B)" means a bucket includes 2 data blocks.
 321
 322----------------------
 323A : bucket
 324B : block
 325N : MAX_DIR_HASH_DEPTH
 326----------------------
 327
 328level #0   | A(2B)
 329           |
 330level #1   | A(2B) - A(2B)
 331           |
 332level #2   | A(2B) - A(2B) - A(2B) - A(2B)
 333     .     |   .       .       .       .
 334level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B)
 335     .     |   .       .       .       .
 336level #N   | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B)
 337
 338The number of blocks and buckets are determined by,
 339
 340                            ,- 2, if n < MAX_DIR_HASH_DEPTH / 2,
 341  # of blocks in level #n = |
 342                            `- 4, Otherwise
 343
 344                             ,- 2^n, if n < MAX_DIR_HASH_DEPTH / 2,
 345  # of buckets in level #n = |
 346                             `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1), Otherwise
 347
 348When F2FS finds a file name in a directory, at first a hash value of the file
 349name is calculated. Then, F2FS scans the hash table in level #0 to find the
 350dentry consisting of the file name and its inode number. If not found, F2FS
 351scans the next hash table in level #1. In this way, F2FS scans hash tables in
 352each levels incrementally from 1 to N. In each levels F2FS needs to scan only
 353one bucket determined by the following equation, which shows O(log(# of files))
 354complexity.
 355
 356  bucket number to scan in level #n = (hash value) % (# of buckets in level #n)
 357
 358In the case of file creation, F2FS finds empty consecutive slots that cover the
 359file name. F2FS searches the empty slots in the hash tables of whole levels from
 3601 to N in the same way as the lookup operation.
 361
 362The following figure shows an example of two cases holding children.
 363       --------------> Dir <--------------
 364       |                                 |
 365    child                             child
 366
 367    child - child                     [hole] - child
 368
 369    child - child - child             [hole] - [hole] - child
 370
 371   Case 1:                           Case 2:
 372   Number of children = 6,           Number of children = 3,
 373   File size = 7                     File size = 7
 374
 375Default Block Allocation
 376------------------------
 377
 378At runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node
 379and Hot/Warm/Cold data.
 380
 381- Hot node      contains direct node blocks of directories.
 382- Warm node     contains direct node blocks except hot node blocks.
 383- Cold node     contains indirect node blocks
 384- Hot data      contains dentry blocks
 385- Warm data     contains data blocks except hot and cold data blocks
 386- Cold data     contains multimedia data or migrated data blocks
 387
 388LFS has two schemes for free space management: threaded log and copy-and-compac-
 389tion. The copy-and-compaction scheme which is known as cleaning, is well-suited
 390for devices showing very good sequential write performance, since free segments
 391are served all the time for writing new data. However, it suffers from cleaning
 392overhead under high utilization. Contrarily, the threaded log scheme suffers
 393from random writes, but no cleaning process is needed. F2FS adopts a hybrid
 394scheme where the copy-and-compaction scheme is adopted by default, but the
 395policy is dynamically changed to the threaded log scheme according to the file
 396system status.
 397
 398In order to align F2FS with underlying flash-based storage, F2FS allocates a
 399segment in a unit of section. F2FS expects that the section size would be the
 400same as the unit size of garbage collection in FTL. Furthermore, with respect
 401to the mapping granularity in FTL, F2FS allocates each section of the active
 402logs from different zones as much as possible, since FTL can write the data in
 403the active logs into one allocation unit according to its mapping granularity.
 404
 405Cleaning process
 406----------------
 407
 408F2FS does cleaning both on demand and in the background. On-demand cleaning is
 409triggered when there are not enough free segments to serve VFS calls. Background
 410cleaner is operated by a kernel thread, and triggers the cleaning job when the
 411system is idle.
 412
 413F2FS supports two victim selection policies: greedy and cost-benefit algorithms.
 414In the greedy algorithm, F2FS selects a victim segment having the smallest number
 415of valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment
 416according to the segment age and the number of valid blocks in order to address
 417log block thrashing problem in the greedy algorithm. F2FS adopts the greedy
 418algorithm for on-demand cleaner, while background cleaner adopts cost-benefit
 419algorithm.
 420
 421In order to identify whether the data in the victim segment are valid or not,
 422F2FS manages a bitmap. Each bit represents the validity of a block, and the
 423bitmap is composed of a bit stream covering whole blocks in main area.
 424