| blob | 13c2ff0343485f7595b66da41aa96475a4602bfd |
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
42 10, 1, 26–52.
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/
56 Cleaning Overhead
57 -----------------
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 ================================================================================
75 KEY FEATURES
76 ================================================================================
78 Flash 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
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.
90 Cleaning 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
97 ================================================================================
98 MOUNT OPTIONS
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
114 segment is cleaned.
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.
124 Default number is 6.
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 noinline_xattr Disable the inline xattrs feature.
129 inline_data Enable the inline data feature: New created small(<~3.4k)
130 files can be written into inode block.
131 inline_dentry Enable the inline dir feature: data in new created
132 directory entries can be written into inode block. The
133 space of inode block which is used to store inline
134 dentries is limited to ~3.4k.
135 noinline_dentry Disable the inline dentry feature.
136 flush_merge Merge concurrent cache_flush commands as much as possible
137 to eliminate redundant command issues. If the underlying
138 device handles the cache_flush command relatively slowly,
139 recommend to enable this option.
140 nobarrier This option can be used if underlying storage guarantees
141 its cached data should be written to the novolatile area.
142 If this option is set, no cache_flush commands are issued
143 but f2fs still guarantees the write ordering of all the
144 data writes.
145 fastboot This option is used when a system wants to reduce mount
146 time as much as possible, even though normal performance
147 can be sacrificed.
148 extent_cache Enable an extent cache based on rb-tree, it can cache
149 as many as extent which map between contiguous logical
150 address and physical address per inode, resulting in
151 increasing the cache hit ratio. Set by default.
152 noextent_cache Disable an extent cache based on rb-tree explicitly, see
153 the above extent_cache mount option.
154 noinline_data Disable the inline data feature, inline data feature is
155 enabled by default.
156 data_flush Enable data flushing before checkpoint in order to
157 persist data of regular and symlink.
158 fault_injection=%d Enable fault injection in all supported types with
159 specified injection rate.
160 mode=%s Control block allocation mode which supports "adaptive"
161 and "lfs". In "lfs" mode, there should be no random
162 writes towards main area.
163 io_bits=%u Set the bit size of write IO requests. It should be set
164 with "mode=lfs".
165 usrquota Enable plain user disk quota accounting.
166 grpquota Enable plain group disk quota accounting.
167 prjquota Enable plain project quota accounting.
168 usrjquota=<file> Appoint specified file and type during mount, so that quota
169 grpjquota=<file> information can be properly updated during recovery flow,
170 prjjquota=<file> <quota file>: must be in root directory;
171 jqfmt=<quota type> <quota type>: [vfsold,vfsv0,vfsv1].
172 offusrjquota Turn off user journelled quota.
173 offgrpjquota Turn off group journelled quota.
174 offprjjquota Turn off project journelled quota.
175 quota Enable plain user disk quota accounting.
176 noquota Disable all plain disk quota option.
178 ================================================================================
179 DEBUGFS ENTRIES
180 ================================================================================
182 /sys/kernel/debug/f2fs/ contains information about all the partitions mounted as
183 f2fs. Each file shows the whole f2fs information.
185 /sys/kernel/debug/f2fs/status includes:
186 - major file system information managed by f2fs currently
187 - average SIT information about whole segments
188 - current memory footprint consumed by f2fs.
190 ================================================================================
191 SYSFS ENTRIES
192 ================================================================================
194 Information about mounted f2fs file systems can be found in
195 /sys/fs/f2fs. Each mounted filesystem will have a directory in
196 /sys/fs/f2fs based on its device name (i.e., /sys/fs/f2fs/sda).
197 The files in each per-device directory are shown in table below.
199 Files in /sys/fs/f2fs/<devname>
200 (see also Documentation/ABI/testing/sysfs-fs-f2fs)
201 ..............................................................................
202 File Content
204 gc_max_sleep_time This tuning parameter controls the maximum sleep
205 time for the garbage collection thread. Time is
206 in milliseconds.
208 gc_min_sleep_time This tuning parameter controls the minimum sleep
209 time for the garbage collection thread. Time is
210 in milliseconds.
212 gc_no_gc_sleep_time This tuning parameter controls the default sleep
213 time for the garbage collection thread. Time is
214 in milliseconds.
216 gc_idle This parameter controls the selection of victim
217 policy for garbage collection. Setting gc_idle = 0
218 (default) will disable this option. Setting
219 gc_idle = 1 will select the Cost Benefit approach
220 & setting gc_idle = 2 will select the greedy approach.
222 gc_urgent This parameter controls triggering background GCs
223 urgently or not. Setting gc_urgent = 0 [default]
224 makes back to default behavior, while if it is set
225 to 1, background thread starts to do GC by given
226 gc_urgent_sleep_time interval.
228 gc_urgent_sleep_time This parameter controls sleep time for gc_urgent.
229 500 ms is set by default. See above gc_urgent.
231 reclaim_segments This parameter controls the number of prefree
232 segments to be reclaimed. If the number of prefree
233 segments is larger than the number of segments
234 in the proportion to the percentage over total
235 volume size, f2fs tries to conduct checkpoint to
236 reclaim the prefree segments to free segments.
237 By default, 5% over total # of segments.
239 max_small_discards This parameter controls the number of discard
240 commands that consist small blocks less than 2MB.
241 The candidates to be discarded are cached until
242 checkpoint is triggered, and issued during the
243 checkpoint. By default, it is disabled with 0.
245 trim_sections This parameter controls the number of sections
246 to be trimmed out in batch mode when FITRIM
247 conducts. 32 sections is set by default.
249 ipu_policy This parameter controls the policy of in-place
250 updates in f2fs. There are five policies:
251 0x01: F2FS_IPU_FORCE, 0x02: F2FS_IPU_SSR,
252 0x04: F2FS_IPU_UTIL, 0x08: F2FS_IPU_SSR_UTIL,
253 0x10: F2FS_IPU_FSYNC.
255 min_ipu_util This parameter controls the threshold to trigger
256 in-place-updates. The number indicates percentage
257 of the filesystem utilization, and used by
258 F2FS_IPU_UTIL and F2FS_IPU_SSR_UTIL policies.
260 min_fsync_blocks This parameter controls the threshold to trigger
261 in-place-updates when F2FS_IPU_FSYNC mode is set.
262 The number indicates the number of dirty pages
263 when fsync needs to flush on its call path. If
264 the number is less than this value, it triggers
265 in-place-updates.
267 max_victim_search This parameter controls the number of trials to
268 find a victim segment when conducting SSR and
269 cleaning operations. The default value is 4096
270 which covers 8GB block address range.
272 dir_level This parameter controls the directory level to
273 support large directory. If a directory has a
274 number of files, it can reduce the file lookup
275 latency by increasing this dir_level value.
276 Otherwise, it needs to decrease this value to
277 reduce the space overhead. The default value is 0.
279 ram_thresh This parameter controls the memory footprint used
280 by free nids and cached nat entries. By default,
281 10 is set, which indicates 10 MB / 1 GB RAM.
283 ================================================================================
284 USAGE
285 ================================================================================
287 1. Download userland tools and compile them.
289 2. Skip, if f2fs was compiled statically inside kernel.
290 Otherwise, insert the f2fs.ko module.
291 # insmod f2fs.ko
293 3. Create a directory trying to mount
294 # mkdir /mnt/f2fs
296 4. Format the block device, and then mount as f2fs
297 # mkfs.f2fs -l label /dev/block_device
298 # mount -t f2fs /dev/block_device /mnt/f2fs
300 mkfs.f2fs
301 ---------
302 The mkfs.f2fs is for the use of formatting a partition as the f2fs filesystem,
303 which builds a basic on-disk layout.
305 The options consist of:
306 -l [label] : Give a volume label, up to 512 unicode name.
307 -a [0 or 1] : Split start location of each area for heap-based allocation.
308 1 is set by default, which performs this.
309 -o [int] : Set overprovision ratio in percent over volume size.
310 5 is set by default.
311 -s [int] : Set the number of segments per section.
312 1 is set by default.
313 -z [int] : Set the number of sections per zone.
314 1 is set by default.
315 -e [str] : Set basic extension list. e.g. "mp3,gif,mov"
316 -t [0 or 1] : Disable discard command or not.
317 1 is set by default, which conducts discard.
319 fsck.f2fs
320 ---------
321 The fsck.f2fs is a tool to check the consistency of an f2fs-formatted
322 partition, which examines whether the filesystem metadata and user-made data
323 are cross-referenced correctly or not.
324 Note that, initial version of the tool does not fix any inconsistency.
326 The options consist of:
327 -d debug level [default:0]
329 dump.f2fs
330 ---------
331 The dump.f2fs shows the information of specific inode and dumps SSA and SIT to
332 file. Each file is dump_ssa and dump_sit.
334 The dump.f2fs is used to debug on-disk data structures of the f2fs filesystem.
335 It shows on-disk inode information recognized by a given inode number, and is
336 able to dump all the SSA and SIT entries into predefined files, ./dump_ssa and
337 ./dump_sit respectively.
339 The options consist of:
340 -d debug level [default:0]
341 -i inode no (hex)
342 -s [SIT dump segno from #1~#2 (decimal), for all 0~-1]
343 -a [SSA dump segno from #1~#2 (decimal), for all 0~-1]
345 Examples:
346 # dump.f2fs -i [ino] /dev/sdx
347 # dump.f2fs -s 0~-1 /dev/sdx (SIT dump)
348 # dump.f2fs -a 0~-1 /dev/sdx (SSA dump)
350 ================================================================================
351 DESIGN
352 ================================================================================
354 On-disk Layout
355 --------------
357 F2FS divides the whole volume into a number of segments, each of which is fixed
358 to 2MB in size. A section is composed of consecutive segments, and a zone
359 consists of a set of sections. By default, section and zone sizes are set to one
360 segment size identically, but users can easily modify the sizes by mkfs.
362 F2FS splits the entire volume into six areas, and all the areas except superblock
363 consists of multiple segments as described below.
365 align with the zone size <-|
366 |-> align with the segment size
367 _________________________________________________________________________
368 | | | Segment | Node | Segment | |
369 | Superblock | Checkpoint | Info. | Address | Summary | Main |
370 | (SB) | (CP) | Table (SIT) | Table (NAT) | Area (SSA) | |
371 |____________|_____2______|______N______|______N______|______N_____|__N___|
372 . .
373 . .
374 . .
375 ._________________________________________.
376 |_Segment_|_..._|_Segment_|_..._|_Segment_|
377 . .
378 ._________._________
379 |_section_|__...__|_
380 . .
381 .________.
382 |__zone__|
384 - Superblock (SB)
385 : It is located at the beginning of the partition, and there exist two copies
386 to avoid file system crash. It contains basic partition information and some
387 default parameters of f2fs.
389 - Checkpoint (CP)
390 : It contains file system information, bitmaps for valid NAT/SIT sets, orphan
391 inode lists, and summary entries of current active segments.
393 - Segment Information Table (SIT)
394 : It contains segment information such as valid block count and bitmap for the
395 validity of all the blocks.
397 - Node Address Table (NAT)
398 : It is composed of a block address table for all the node blocks stored in
399 Main area.
401 - Segment Summary Area (SSA)
402 : It contains summary entries which contains the owner information of all the
403 data and node blocks stored in Main area.
405 - Main Area
406 : It contains file and directory data including their indices.
408 In order to avoid misalignment between file system and flash-based storage, F2FS
409 aligns the start block address of CP with the segment size. Also, it aligns the
410 start block address of Main area with the zone size by reserving some segments
411 in SSA area.
413 Reference the following survey for additional technical details.
414 https://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey
416 File System Metadata Structure
417 ------------------------------
419 F2FS adopts the checkpointing scheme to maintain file system consistency. At
420 mount time, F2FS first tries to find the last valid checkpoint data by scanning
421 CP area. In order to reduce the scanning time, F2FS uses only two copies of CP.
422 One of them always indicates the last valid data, which is called as shadow copy
423 mechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism.
425 For file system consistency, each CP points to which NAT and SIT copies are
426 valid, as shown as below.
428 +--------+----------+---------+
429 | CP | SIT | NAT |
430 +--------+----------+---------+
431 . . . .
432 . . . .
433 . . . .
434 +-------+-------+--------+--------+--------+--------+
435 | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 |
436 +-------+-------+--------+--------+--------+--------+
437 | ^ ^
438 | | |
439 `----------------------------------------'
441 Index Structure
442 ---------------
444 The key data structure to manage the data locations is a "node". Similar to
445 traditional file structures, F2FS has three types of node: inode, direct node,
446 indirect node. F2FS assigns 4KB to an inode block which contains 923 data block
447 indices, two direct node pointers, two indirect node pointers, and one double
448 indirect node pointer as described below. One direct node block contains 1018
449 data blocks, and one indirect node block contains also 1018 node blocks. Thus,
450 one inode block (i.e., a file) covers:
452 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB.
454 Inode block (4KB)
455 |- data (923)
456 |- direct node (2)
457 | `- data (1018)
458 |- indirect node (2)
459 | `- direct node (1018)
460 | `- data (1018)
461 `- double indirect node (1)
462 `- indirect node (1018)
463 `- direct node (1018)
464 `- data (1018)
466 Note that, all the node blocks are mapped by NAT which means the location of
467 each node is translated by the NAT table. In the consideration of the wandering
468 tree problem, F2FS is able to cut off the propagation of node updates caused by
469 leaf data writes.
471 Directory Structure
472 -------------------
474 A directory entry occupies 11 bytes, which consists of the following attributes.
476 - hash hash value of the file name
477 - ino inode number
478 - len the length of file name
479 - type file type such as directory, symlink, etc
481 A dentry block consists of 214 dentry slots and file names. Therein a bitmap is
482 used to represent whether each dentry is valid or not. A dentry block occupies
483 4KB with the following composition.
485 Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) +
486 dentries(11 * 214 bytes) + file name (8 * 214 bytes)
488 [Bucket]
489 +--------------------------------+
490 |dentry block 1 | dentry block 2 |
491 +--------------------------------+
492 . .
493 . .
494 . [Dentry Block Structure: 4KB] .
495 +--------+----------+----------+------------+
496 | bitmap | reserved | dentries | file names |
497 +--------+----------+----------+------------+
498 [Dentry Block: 4KB] . .
499 . .
500 . .
501 +------+------+-----+------+
502 | hash | ino | len | type |
503 +------+------+-----+------+
504 [Dentry Structure: 11 bytes]
506 F2FS implements multi-level hash tables for directory structure. Each level has
507 a hash table with dedicated number of hash buckets as shown below. Note that
508 "A(2B)" means a bucket includes 2 data blocks.
510 ----------------------
511 A : bucket
512 B : block
513 N : MAX_DIR_HASH_DEPTH
514 ----------------------
516 level #0 | A(2B)
517 |
518 level #1 | A(2B) - A(2B)
519 |
520 level #2 | A(2B) - A(2B) - A(2B) - A(2B)
521 . | . . . .
522 level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B)
523 . | . . . .
524 level #N | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B)
526 The number of blocks and buckets are determined by,
528 ,- 2, if n < MAX_DIR_HASH_DEPTH / 2,
529 # of blocks in level #n = |
530 `- 4, Otherwise
532 ,- 2^(n + dir_level),
533 | if n + dir_level < MAX_DIR_HASH_DEPTH / 2,
534 # of buckets in level #n = |
535 `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1),
536 Otherwise
538 When F2FS finds a file name in a directory, at first a hash value of the file
539 name is calculated. Then, F2FS scans the hash table in level #0 to find the
540 dentry consisting of the file name and its inode number. If not found, F2FS
541 scans the next hash table in level #1. In this way, F2FS scans hash tables in
542 each levels incrementally from 1 to N. In each levels F2FS needs to scan only
543 one bucket determined by the following equation, which shows O(log(# of files))
544 complexity.
546 bucket number to scan in level #n = (hash value) % (# of buckets in level #n)
548 In the case of file creation, F2FS finds empty consecutive slots that cover the
549 file name. F2FS searches the empty slots in the hash tables of whole levels from
550 1 to N in the same way as the lookup operation.
552 The following figure shows an example of two cases holding children.
553 --------------> Dir <--------------
554 | |
555 child child
557 child - child [hole] - child
559 child - child - child [hole] - [hole] - child
561 Case 1: Case 2:
562 Number of children = 6, Number of children = 3,
563 File size = 7 File size = 7
565 Default Block Allocation
566 ------------------------
568 At runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node
569 and Hot/Warm/Cold data.
571 - Hot node contains direct node blocks of directories.
572 - Warm node contains direct node blocks except hot node blocks.
573 - Cold node contains indirect node blocks
574 - Hot data contains dentry blocks
575 - Warm data contains data blocks except hot and cold data blocks
576 - Cold data contains multimedia data or migrated data blocks
578 LFS has two schemes for free space management: threaded log and copy-and-compac-
579 tion. The copy-and-compaction scheme which is known as cleaning, is well-suited
580 for devices showing very good sequential write performance, since free segments
581 are served all the time for writing new data. However, it suffers from cleaning
582 overhead under high utilization. Contrarily, the threaded log scheme suffers
583 from random writes, but no cleaning process is needed. F2FS adopts a hybrid
584 scheme where the copy-and-compaction scheme is adopted by default, but the
585 policy is dynamically changed to the threaded log scheme according to the file
586 system status.
588 In order to align F2FS with underlying flash-based storage, F2FS allocates a
589 segment in a unit of section. F2FS expects that the section size would be the
590 same as the unit size of garbage collection in FTL. Furthermore, with respect
591 to the mapping granularity in FTL, F2FS allocates each section of the active
592 logs from different zones as much as possible, since FTL can write the data in
593 the active logs into one allocation unit according to its mapping granularity.
595 Cleaning process
596 ----------------
598 F2FS does cleaning both on demand and in the background. On-demand cleaning is
599 triggered when there are not enough free segments to serve VFS calls. Background
600 cleaner is operated by a kernel thread, and triggers the cleaning job when the
601 system is idle.
603 F2FS supports two victim selection policies: greedy and cost-benefit algorithms.
604 In the greedy algorithm, F2FS selects a victim segment having the smallest number
605 of valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment
606 according to the segment age and the number of valid blocks in order to address
607 log block thrashing problem in the greedy algorithm. F2FS adopts the greedy
608 algorithm for on-demand cleaner, while background cleaner adopts cost-benefit
609 algorithm.
611 In order to identify whether the data in the victim segment are valid or not,
612 F2FS manages a bitmap. Each bit represents the validity of a block, and the
613 bitmap is composed of a bit stream covering whole blocks in main area.