1 ================================================================================
2 WHAT IS Flash-Friendly File System (F2FS)?
3 ================================================================================
5 NAND flash memory-based storage devices, such as SSD, eMMC, and SD cards, have
6 been equipped on a variety systems ranging from mobile to server systems. Since
7 they are known to have different characteristics from the conventional rotating
8 disks, a file system, an upper layer to the storage device, should adapt to the
9 changes from the sketch in the design level.
11 F2FS is a file system exploiting NAND flash memory-based storage devices, which
12 is based on Log-structured File System (LFS). The design has been focused on
13 addressing the fundamental issues in LFS, which are snowball effect of wandering
14 tree and high cleaning overhead.
16 Since a NAND flash memory-based storage device shows different characteristic
17 according to its internal geometry or flash memory management scheme, namely FTL,
18 F2FS and its tools support various parameters not only for configuring on-disk
19 layout, but also for selecting allocation and cleaning algorithms.
21 The following git tree provides the file system formatting tool (mkfs.f2fs),
22 a consistency checking tool (fsck.f2fs), and a debugging tool (dump.f2fs).
23 >> git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs-tools.git
25 For reporting bugs and sending patches, please use the following mailing list:
26 >> linux-f2fs-devel@lists.sourceforge.net
28 ================================================================================
29 BACKGROUND AND DESIGN ISSUES
30 ================================================================================
32 Log-structured File System (LFS)
33 --------------------------------
34 "A log-structured file system writes all modifications to disk sequentially in
35 a log-like structure, thereby speeding up both file writing and crash recovery.
36 The log is the only structure on disk; it contains indexing information so that
37 files can be read back from the log efficiently. In order to maintain large free
38 areas on disk for fast writing, we divide the log into segments and use a
39 segment cleaner to compress the live information from heavily fragmented
40 segments." from Rosenblum, M. and Ousterhout, J. K., 1992, "The design and
41 implementation of a log-structured file system", ACM Trans. Computer Systems
44 Wandering Tree Problem
45 ----------------------
46 In LFS, when a file data is updated and written to the end of log, its direct
47 pointer block is updated due to the changed location. Then the indirect pointer
48 block is also updated due to the direct pointer block update. In this manner,
49 the upper index structures such as inode, inode map, and checkpoint block are
50 also updated recursively. This problem is called as wandering tree problem [1],
51 and in order to enhance the performance, it should eliminate or relax the update
52 propagation as much as possible.
54 [1] Bityutskiy, A. 2005. JFFS3 design issues. http://www.linux-mtd.infradead.org/
58 Since LFS is based on out-of-place writes, it produces so many obsolete blocks
59 scattered across the whole storage. In order to serve new empty log space, it
60 needs to reclaim these obsolete blocks seamlessly to users. This job is called
61 as a cleaning process.
63 The process consists of three operations as follows.
64 1. A victim segment is selected through referencing segment usage table.
65 2. It loads parent index structures of all the data in the victim identified by
66 segment summary blocks.
67 3. It checks the cross-reference between the data and its parent index structure.
68 4. It moves valid data selectively.
70 This cleaning job may cause unexpected long delays, so the most important goal
71 is to hide the latencies to users. And also definitely, it should reduce the
72 amount of valid data to be moved, and move them quickly as well.
74 ================================================================================
76 ================================================================================
80 - Enlarge the random write area for better performance, but provide the high
82 - Align FS data structures to the operational units in FTL as best efforts
84 Wandering Tree Problem
85 ----------------------
86 - Use a term, “node”, that represents inodes as well as various pointer blocks
87 - Introduce Node Address Table (NAT) containing the locations of all the “node”
88 blocks; this will cut off the update propagation.
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
97 ================================================================================
99 ================================================================================
101 background_gc=%s Turn on/off cleaning operations, namely garbage
102 collection, triggered in background when I/O subsystem is
103 idle. If background_gc=on, it will turn on the garbage
104 collection and if background_gc=off, garbage collection
105 will be turned off. If background_gc=sync, it will turn
106 on synchronous garbage collection running in background.
107 Default value for this option is on. So garbage
108 collection is on by default.
109 disable_roll_forward Disable the roll-forward recovery routine
110 norecovery Disable the roll-forward recovery routine, mounted read-
111 only (i.e., -o ro,disable_roll_forward)
112 discard/nodiscard Enable/disable real-time discard in f2fs, if discard is
113 enabled, f2fs will issue discard/TRIM commands when a
115 no_heap Disable heap-style segment allocation which finds free
116 segments for data from the beginning of main area, while
117 for node from the end of main area.
118 nouser_xattr Disable Extended User Attributes. Note: xattr is enabled
119 by default if CONFIG_F2FS_FS_XATTR is selected.
120 noacl Disable POSIX Access Control List. Note: acl is enabled
121 by default if CONFIG_F2FS_FS_POSIX_ACL is selected.
122 active_logs=%u Support configuring the number of active logs. In the
123 current design, f2fs supports only 2, 4, and 6 logs.
125 disable_ext_identify Disable the extension list configured by mkfs, so f2fs
126 does not aware of cold files such as media files.
127 inline_xattr Enable the inline xattrs feature.
128 inline_data Enable the inline data feature: New created small(<~3.4k)
129 files can be written into inode block.
130 inline_dentry Enable the inline dir feature: data in new created
131 directory entries can be written into inode block. The
132 space of inode block which is used to store inline
133 dentries is limited to ~3.4k.
134 flush_merge Merge concurrent cache_flush commands as much as possible
135 to eliminate redundant command issues. If the underlying
136 device handles the cache_flush command relatively slowly,
137 recommend to enable this option.
138 nobarrier This option can be used if underlying storage guarantees
139 its cached data should be written to the novolatile area.
140 If this option is set, no cache_flush commands are issued
141 but f2fs still guarantees the write ordering of all the
143 fastboot This option is used when a system wants to reduce mount
144 time as much as possible, even though normal performance
146 extent_cache Enable an extent cache based on rb-tree, it can cache
147 as many as extent which map between contiguous logical
148 address and physical address per inode, resulting in
149 increasing the cache hit ratio. Set by default.
150 noextent_cache Disable an extent cache based on rb-tree explicitly, see
151 the above extent_cache mount option.
152 noinline_data Disable the inline data feature, inline data feature is
154 data_flush Enable data flushing before checkpoint in order to
155 persist data of regular and symlink.
156 mode=%s Control block allocation mode which supports "adaptive"
157 and "lfs". In "lfs" mode, there should be no random
158 writes towards main area.
160 ================================================================================
162 ================================================================================
164 /sys/kernel/debug/f2fs/ contains information about all the partitions mounted as
165 f2fs. Each file shows the whole f2fs information.
167 /sys/kernel/debug/f2fs/status includes:
168 - major file system information managed by f2fs currently
169 - average SIT information about whole segments
170 - current memory footprint consumed by f2fs.
172 ================================================================================
174 ================================================================================
176 Information about mounted f2f2 file systems can be found in
177 /sys/fs/f2fs. Each mounted filesystem will have a directory in
178 /sys/fs/f2fs based on its device name (i.e., /sys/fs/f2fs/sda).
179 The files in each per-device directory are shown in table below.
181 Files in /sys/fs/f2fs/<devname>
182 (see also Documentation/ABI/testing/sysfs-fs-f2fs)
183 ..............................................................................
186 gc_max_sleep_time This tuning parameter controls the maximum sleep
187 time for the garbage collection thread. Time is
190 gc_min_sleep_time This tuning parameter controls the minimum sleep
191 time for the garbage collection thread. Time is
194 gc_no_gc_sleep_time This tuning parameter controls the default sleep
195 time for the garbage collection thread. Time is
198 gc_idle This parameter controls the selection of victim
199 policy for garbage collection. Setting gc_idle = 0
200 (default) will disable this option. Setting
201 gc_idle = 1 will select the Cost Benefit approach
202 & setting gc_idle = 2 will select the greedy approach.
204 reclaim_segments This parameter controls the number of prefree
205 segments to be reclaimed. If the number of prefree
206 segments is larger than the number of segments
207 in the proportion to the percentage over total
208 volume size, f2fs tries to conduct checkpoint to
209 reclaim the prefree segments to free segments.
210 By default, 5% over total # of segments.
212 max_small_discards This parameter controls the number of discard
213 commands that consist small blocks less than 2MB.
214 The candidates to be discarded are cached until
215 checkpoint is triggered, and issued during the
216 checkpoint. By default, it is disabled with 0.
218 trim_sections This parameter controls the number of sections
219 to be trimmed out in batch mode when FITRIM
220 conducts. 32 sections is set by default.
222 ipu_policy This parameter controls the policy of in-place
223 updates in f2fs. There are five policies:
224 0x01: F2FS_IPU_FORCE, 0x02: F2FS_IPU_SSR,
225 0x04: F2FS_IPU_UTIL, 0x08: F2FS_IPU_SSR_UTIL,
226 0x10: F2FS_IPU_FSYNC.
228 min_ipu_util This parameter controls the threshold to trigger
229 in-place-updates. The number indicates percentage
230 of the filesystem utilization, and used by
231 F2FS_IPU_UTIL and F2FS_IPU_SSR_UTIL policies.
233 min_fsync_blocks This parameter controls the threshold to trigger
234 in-place-updates when F2FS_IPU_FSYNC mode is set.
235 The number indicates the number of dirty pages
236 when fsync needs to flush on its call path. If
237 the number is less than this value, it triggers
240 max_victim_search This parameter controls the number of trials to
241 find a victim segment when conducting SSR and
242 cleaning operations. The default value is 4096
243 which covers 8GB block address range.
245 dir_level This parameter controls the directory level to
246 support large directory. If a directory has a
247 number of files, it can reduce the file lookup
248 latency by increasing this dir_level value.
249 Otherwise, it needs to decrease this value to
250 reduce the space overhead. The default value is 0.
252 ram_thresh This parameter controls the memory footprint used
253 by free nids and cached nat entries. By default,
254 10 is set, which indicates 10 MB / 1 GB RAM.
256 ================================================================================
258 ================================================================================
260 1. Download userland tools and compile them.
262 2. Skip, if f2fs was compiled statically inside kernel.
263 Otherwise, insert the f2fs.ko module.
266 3. Create a directory trying to mount
269 4. Format the block device, and then mount as f2fs
270 # mkfs.f2fs -l label /dev/block_device
271 # mount -t f2fs /dev/block_device /mnt/f2fs
275 The mkfs.f2fs is for the use of formatting a partition as the f2fs filesystem,
276 which builds a basic on-disk layout.
278 The options consist of:
279 -l [label] : Give a volume label, up to 512 unicode name.
280 -a [0 or 1] : Split start location of each area for heap-based allocation.
281 1 is set by default, which performs this.
282 -o [int] : Set overprovision ratio in percent over volume size.
284 -s [int] : Set the number of segments per section.
286 -z [int] : Set the number of sections per zone.
288 -e [str] : Set basic extension list. e.g. "mp3,gif,mov"
289 -t [0 or 1] : Disable discard command or not.
290 1 is set by default, which conducts discard.
294 The fsck.f2fs is a tool to check the consistency of an f2fs-formatted
295 partition, which examines whether the filesystem metadata and user-made data
296 are cross-referenced correctly or not.
297 Note that, initial version of the tool does not fix any inconsistency.
299 The options consist of:
300 -d debug level [default:0]
304 The dump.f2fs shows the information of specific inode and dumps SSA and SIT to
305 file. Each file is dump_ssa and dump_sit.
307 The dump.f2fs is used to debug on-disk data structures of the f2fs filesystem.
308 It shows on-disk inode information recognized by a given inode number, and is
309 able to dump all the SSA and SIT entries into predefined files, ./dump_ssa and
310 ./dump_sit respectively.
312 The options consist of:
313 -d debug level [default:0]
315 -s [SIT dump segno from #1~#2 (decimal), for all 0~-1]
316 -a [SSA dump segno from #1~#2 (decimal), for all 0~-1]
319 # dump.f2fs -i [ino] /dev/sdx
320 # dump.f2fs -s 0~-1 /dev/sdx (SIT dump)
321 # dump.f2fs -a 0~-1 /dev/sdx (SSA dump)
323 ================================================================================
325 ================================================================================
330 F2FS divides the whole volume into a number of segments, each of which is fixed
331 to 2MB in size. A section is composed of consecutive segments, and a zone
332 consists of a set of sections. By default, section and zone sizes are set to one
333 segment size identically, but users can easily modify the sizes by mkfs.
335 F2FS splits the entire volume into six areas, and all the areas except superblock
336 consists of multiple segments as described below.
338 align with the zone size <-|
339 |-> align with the segment size
340 _________________________________________________________________________
341 | | | Segment | Node | Segment | |
342 | Superblock | Checkpoint | Info. | Address | Summary | Main |
343 | (SB) | (CP) | Table (SIT) | Table (NAT) | Area (SSA) | |
344 |____________|_____2______|______N______|______N______|______N_____|__N___|
348 ._________________________________________.
349 |_Segment_|_..._|_Segment_|_..._|_Segment_|
358 : It is located at the beginning of the partition, and there exist two copies
359 to avoid file system crash. It contains basic partition information and some
360 default parameters of f2fs.
363 : It contains file system information, bitmaps for valid NAT/SIT sets, orphan
364 inode lists, and summary entries of current active segments.
366 - Segment Information Table (SIT)
367 : It contains segment information such as valid block count and bitmap for the
368 validity of all the blocks.
370 - Node Address Table (NAT)
371 : It is composed of a block address table for all the node blocks stored in
374 - Segment Summary Area (SSA)
375 : It contains summary entries which contains the owner information of all the
376 data and node blocks stored in Main area.
379 : It contains file and directory data including their indices.
381 In order to avoid misalignment between file system and flash-based storage, F2FS
382 aligns the start block address of CP with the segment size. Also, it aligns the
383 start block address of Main area with the zone size by reserving some segments
386 Reference the following survey for additional technical details.
387 https://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey
389 File System Metadata Structure
390 ------------------------------
392 F2FS adopts the checkpointing scheme to maintain file system consistency. At
393 mount time, F2FS first tries to find the last valid checkpoint data by scanning
394 CP area. In order to reduce the scanning time, F2FS uses only two copies of CP.
395 One of them always indicates the last valid data, which is called as shadow copy
396 mechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism.
398 For file system consistency, each CP points to which NAT and SIT copies are
399 valid, as shown as below.
401 +--------+----------+---------+
403 +--------+----------+---------+
407 +-------+-------+--------+--------+--------+--------+
408 | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 |
409 +-------+-------+--------+--------+--------+--------+
412 `----------------------------------------'
417 The key data structure to manage the data locations is a "node". Similar to
418 traditional file structures, F2FS has three types of node: inode, direct node,
419 indirect node. F2FS assigns 4KB to an inode block which contains 923 data block
420 indices, two direct node pointers, two indirect node pointers, and one double
421 indirect node pointer as described below. One direct node block contains 1018
422 data blocks, and one indirect node block contains also 1018 node blocks. Thus,
423 one inode block (i.e., a file) covers:
425 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB.
432 | `- direct node (1018)
434 `- double indirect node (1)
435 `- indirect node (1018)
436 `- direct node (1018)
439 Note that, all the node blocks are mapped by NAT which means the location of
440 each node is translated by the NAT table. In the consideration of the wandering
441 tree problem, F2FS is able to cut off the propagation of node updates caused by
447 A directory entry occupies 11 bytes, which consists of the following attributes.
449 - hash hash value of the file name
451 - len the length of file name
452 - type file type such as directory, symlink, etc
454 A dentry block consists of 214 dentry slots and file names. Therein a bitmap is
455 used to represent whether each dentry is valid or not. A dentry block occupies
456 4KB with the following composition.
458 Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) +
459 dentries(11 * 214 bytes) + file name (8 * 214 bytes)
462 +--------------------------------+
463 |dentry block 1 | dentry block 2 |
464 +--------------------------------+
467 . [Dentry Block Structure: 4KB] .
468 +--------+----------+----------+------------+
469 | bitmap | reserved | dentries | file names |
470 +--------+----------+----------+------------+
471 [Dentry Block: 4KB] . .
474 +------+------+-----+------+
475 | hash | ino | len | type |
476 +------+------+-----+------+
477 [Dentry Structure: 11 bytes]
479 F2FS implements multi-level hash tables for directory structure. Each level has
480 a hash table with dedicated number of hash buckets as shown below. Note that
481 "A(2B)" means a bucket includes 2 data blocks.
483 ----------------------
486 N : MAX_DIR_HASH_DEPTH
487 ----------------------
491 level #1 | A(2B) - A(2B)
493 level #2 | A(2B) - A(2B) - A(2B) - A(2B)
495 level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B)
497 level #N | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B)
499 The number of blocks and buckets are determined by,
501 ,- 2, if n < MAX_DIR_HASH_DEPTH / 2,
502 # of blocks in level #n = |
505 ,- 2^(n + dir_level),
506 | if n + dir_level < MAX_DIR_HASH_DEPTH / 2,
507 # of buckets in level #n = |
508 `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1),
511 When F2FS finds a file name in a directory, at first a hash value of the file
512 name is calculated. Then, F2FS scans the hash table in level #0 to find the
513 dentry consisting of the file name and its inode number. If not found, F2FS
514 scans the next hash table in level #1. In this way, F2FS scans hash tables in
515 each levels incrementally from 1 to N. In each levels F2FS needs to scan only
516 one bucket determined by the following equation, which shows O(log(# of files))
519 bucket number to scan in level #n = (hash value) % (# of buckets in level #n)
521 In the case of file creation, F2FS finds empty consecutive slots that cover the
522 file name. F2FS searches the empty slots in the hash tables of whole levels from
523 1 to N in the same way as the lookup operation.
525 The following figure shows an example of two cases holding children.
526 --------------> Dir <--------------
530 child - child [hole] - child
532 child - child - child [hole] - [hole] - child
535 Number of children = 6, Number of children = 3,
536 File size = 7 File size = 7
538 Default Block Allocation
539 ------------------------
541 At runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node
542 and Hot/Warm/Cold data.
544 - Hot node contains direct node blocks of directories.
545 - Warm node contains direct node blocks except hot node blocks.
546 - Cold node contains indirect node blocks
547 - Hot data contains dentry blocks
548 - Warm data contains data blocks except hot and cold data blocks
549 - Cold data contains multimedia data or migrated data blocks
551 LFS has two schemes for free space management: threaded log and copy-and-compac-
552 tion. The copy-and-compaction scheme which is known as cleaning, is well-suited
553 for devices showing very good sequential write performance, since free segments
554 are served all the time for writing new data. However, it suffers from cleaning
555 overhead under high utilization. Contrarily, the threaded log scheme suffers
556 from random writes, but no cleaning process is needed. F2FS adopts a hybrid
557 scheme where the copy-and-compaction scheme is adopted by default, but the
558 policy is dynamically changed to the threaded log scheme according to the file
561 In order to align F2FS with underlying flash-based storage, F2FS allocates a
562 segment in a unit of section. F2FS expects that the section size would be the
563 same as the unit size of garbage collection in FTL. Furthermore, with respect
564 to the mapping granularity in FTL, F2FS allocates each section of the active
565 logs from different zones as much as possible, since FTL can write the data in
566 the active logs into one allocation unit according to its mapping granularity.
571 F2FS does cleaning both on demand and in the background. On-demand cleaning is
572 triggered when there are not enough free segments to serve VFS calls. Background
573 cleaner is operated by a kernel thread, and triggers the cleaning job when the
576 F2FS supports two victim selection policies: greedy and cost-benefit algorithms.
577 In the greedy algorithm, F2FS selects a victim segment having the smallest number
578 of valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment
579 according to the segment age and the number of valid blocks in order to address
580 log block thrashing problem in the greedy algorithm. F2FS adopts the greedy
581 algorithm for on-demand cleaner, while background cleaner adopts cost-benefit
584 In order to identify whether the data in the victim segment are valid or not,
585 F2FS manages a bitmap. Each bit represents the validity of a block, and the
586 bitmap is composed of a bit stream covering whole blocks in main area.
projects / deliverable / linux.git / blob