You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

2197 lines
84 KiB

7 years ago
7 years ago
8 years ago
9 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
7 years ago
8 years ago
8 years ago
8 years ago
8 years ago
9 years ago
9 years ago
9 years ago
9 years ago
8 years ago
9 years ago
8 years ago
9 years ago
8 years ago
9 years ago
8 years ago
9 years ago
8 years ago
9 years ago
8 years ago
9 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
7 years ago
5 years ago
5 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
7 years ago
8 years ago
8 years ago
5 years ago
5 years ago
8 years ago
8 years ago
8 years ago
8 years ago
8 years ago
9 years ago
8 years ago
8 years ago
7 years ago
7 years ago
8 years ago
8 years ago
8 years ago
  1. .\"t
  2. .\" Automatically generated by Pandoc 1.19.2.4
  3. .\"
  4. .TH "mergerfs" "1" "2019\-12\-10" "mergerfs user manual" ""
  5. .hy
  6. .SH NAME
  7. .PP
  8. mergerfs \- a featureful union filesystem
  9. .SH SYNOPSIS
  10. .PP
  11. mergerfs \-o<options> <branches> <mountpoint>
  12. .SH DESCRIPTION
  13. .PP
  14. \f[B]mergerfs\f[] is a union filesystem geared towards simplifying
  15. storage and management of files across numerous commodity storage
  16. devices.
  17. It is similar to \f[B]mhddfs\f[], \f[B]unionfs\f[], and \f[B]aufs\f[].
  18. .SH FEATURES
  19. .IP \[bu] 2
  20. Runs in userspace (FUSE)
  21. .IP \[bu] 2
  22. Configurable behaviors / file placement
  23. .IP \[bu] 2
  24. Support for extended attributes (xattrs)
  25. .IP \[bu] 2
  26. Support for file attributes (chattr)
  27. .IP \[bu] 2
  28. Runtime configurable (via xattrs)
  29. .IP \[bu] 2
  30. Safe to run as root
  31. .IP \[bu] 2
  32. Opportunistic credential caching
  33. .IP \[bu] 2
  34. Works with heterogeneous filesystem types
  35. .IP \[bu] 2
  36. Handling of writes to full drives (transparently move file to drive with
  37. capacity)
  38. .IP \[bu] 2
  39. Handles pool of read\-only and read/write drives
  40. .IP \[bu] 2
  41. Can turn read\-only files into symlinks to underlying file
  42. .IP \[bu] 2
  43. Hard link copy\-on\-write / CoW
  44. .IP \[bu] 2
  45. supports POSIX ACLs
  46. .SH How it works
  47. .PP
  48. mergerfs logically merges multiple paths together.
  49. Think a union of sets.
  50. The file/s or directory/s acted on or presented through mergerfs are
  51. based on the policy chosen for that particular action.
  52. Read more about policies below.
  53. .IP
  54. .nf
  55. \f[C]
  56. A\ \ \ \ \ \ \ \ \ +\ \ \ \ \ \ B\ \ \ \ \ \ \ \ =\ \ \ \ \ \ \ C
  57. /disk1\ \ \ \ \ \ \ \ \ \ \ /disk2\ \ \ \ \ \ \ \ \ \ \ /merged
  58. |\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |
  59. +\-\-\ /dir1\ \ \ \ \ \ \ \ +\-\-\ /dir1\ \ \ \ \ \ \ \ +\-\-\ /dir1
  60. |\ \ \ |\ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ |\ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ |
  61. |\ \ \ +\-\-\ file1\ \ \ \ |\ \ \ +\-\-\ file2\ \ \ \ |\ \ \ +\-\-\ file1
  62. |\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ +\-\-\ file3\ \ \ \ |\ \ \ +\-\-\ file2
  63. +\-\-\ /dir2\ \ \ \ \ \ \ \ |\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ +\-\-\ file3
  64. |\ \ \ |\ \ \ \ \ \ \ \ \ \ \ \ +\-\-\ /dir3\ \ \ \ \ \ \ \ |
  65. |\ \ \ +\-\-\ file4\ \ \ \ \ \ \ \ |\ \ \ \ \ \ \ \ \ \ \ \ +\-\-\ /dir2
  66. |\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ +\-\-\ file5\ \ \ |\ \ \ |
  67. +\-\-\ file6\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ +\-\-\ file4
  68. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |
  69. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ +\-\-\ /dir3
  70. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ |
  71. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |\ \ \ +\-\-\ file5
  72. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ |
  73. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ +\-\-\ file6
  74. \f[]
  75. .fi
  76. .PP
  77. mergerfs does \f[B]not\f[] support the copy\-on\-write (CoW) behavior
  78. found in \f[B]aufs\f[] and \f[B]overlayfs\f[].
  79. You can \f[B]not\f[] mount a read\-only filesystem and write to it.
  80. However, mergerfs will ignore read\-only drives when creating new files
  81. so you can mix read\-write and read\-only drives.
  82. It also does \f[B]not\f[] split data across drives.
  83. It is not RAID0 / striping.
  84. It is simply a union.
  85. .SH OPTIONS
  86. .SS mount options
  87. .IP \[bu] 2
  88. \f[B]allow_other\f[]: A libfuse option which allows users besides the
  89. one which ran mergerfs to see the filesystem.
  90. This is required for most use\-cases.
  91. .IP \[bu] 2
  92. \f[B]minfreespace=SIZE\f[]: The minimum space value used for creation
  93. policies.
  94. Understands \[aq]K\[aq], \[aq]M\[aq], and \[aq]G\[aq] to represent
  95. kilobyte, megabyte, and gigabyte respectively.
  96. (default: 4G)
  97. .IP \[bu] 2
  98. \f[B]moveonenospc=BOOL\f[]: When enabled if a \f[B]write\f[] fails with
  99. \f[B]ENOSPC\f[] or \f[B]EDQUOT\f[] a scan of all drives will be done
  100. looking for the drive with the most free space which is at least the
  101. size of the file plus the amount which failed to write.
  102. An attempt to move the file to that drive will occur (keeping all
  103. metadata possible) and if successful the original is unlinked and the
  104. write retried.
  105. (default: false)
  106. .IP \[bu] 2
  107. \f[B]use_ino\f[]: Causes mergerfs to supply file/directory inodes rather
  108. than libfuse.
  109. While not a default it is recommended it be enabled so that linked files
  110. share the same inode value.
  111. .IP \[bu] 2
  112. \f[B]dropcacheonclose=BOOL\f[]: When a file is requested to be closed
  113. call \f[C]posix_fadvise\f[] on it first to instruct the kernel that we
  114. no longer need the data and it can drop its cache.
  115. Recommended when \f[B]cache.files=partial|full|auto\-full\f[] to limit
  116. double caching.
  117. (default: false)
  118. .IP \[bu] 2
  119. \f[B]symlinkify=BOOL\f[]: When enabled and a file is not writable and
  120. its mtime or ctime is older than \f[B]symlinkify_timeout\f[] files will
  121. be reported as symlinks to the original files.
  122. Please read more below before using.
  123. (default: false)
  124. .IP \[bu] 2
  125. \f[B]symlinkify_timeout=INT\f[]: Time to wait, in seconds, to activate
  126. the \f[B]symlinkify\f[] behavior.
  127. (default: 3600)
  128. .IP \[bu] 2
  129. \f[B]nullrw=BOOL\f[]: Turns reads and writes into no\-ops.
  130. The request will succeed but do nothing.
  131. Useful for benchmarking mergerfs.
  132. (default: false)
  133. .IP \[bu] 2
  134. \f[B]ignorepponrename=BOOL\f[]: Ignore path preserving on rename.
  135. Typically rename and link act differently depending on the policy of
  136. \f[C]create\f[] (read below).
  137. Enabling this will cause rename and link to always use the non\-path
  138. preserving behavior.
  139. This means files, when renamed or linked, will stay on the same drive.
  140. (default: false)
  141. .IP \[bu] 2
  142. \f[B]security_capability=BOOL\f[]: If false return ENOATTR when xattr
  143. security.capability is queried.
  144. (default: true)
  145. .IP \[bu] 2
  146. \f[B]xattr=passthrough|noattr|nosys\f[]: Runtime control of xattrs.
  147. Default is to passthrough xattr requests.
  148. \[aq]noattr\[aq] will short circuit as if nothing exists.
  149. \[aq]nosys\[aq] will respond with ENOSYS as if xattrs are not supported
  150. or disabled.
  151. (default: passthrough)
  152. .IP \[bu] 2
  153. \f[B]link_cow=BOOL\f[]: When enabled if a regular file is opened which
  154. has a link count > 1 it will copy the file to a temporary file and
  155. rename over the original.
  156. Breaking the link and providing a basic copy\-on\-write function similar
  157. to cow\-shell.
  158. (default: false)
  159. .IP \[bu] 2
  160. \f[B]statfs=base|full\f[]: Controls how statfs works.
  161. \[aq]base\[aq] means it will always use all branches in statfs
  162. calculations.
  163. \[aq]full\[aq] is in effect path preserving and only includes drives
  164. where the path exists.
  165. (default: base)
  166. .IP \[bu] 2
  167. \f[B]statfs_ignore=none|ro|nc\f[]: \[aq]ro\[aq] will cause statfs
  168. calculations to ignore available space for branches mounted or tagged as
  169. \[aq]read\-only\[aq] or \[aq]no create\[aq].
  170. \[aq]nc\[aq] will ignore available space for branches tagged as \[aq]no
  171. create\[aq].
  172. (default: none)
  173. .IP \[bu] 2
  174. \f[B]posix_acl=BOOL\f[]: Enable POSIX ACL support (if supported by
  175. kernel and underlying filesystem).
  176. (default: false)
  177. .IP \[bu] 2
  178. \f[B]async_read=BOOL\f[]: Perform reads asynchronously.
  179. If disabled or unavailable the kernel will ensure there is at most one
  180. pending read request per file handle and will attempt to order requests
  181. by offset.
  182. (default: true)
  183. .IP \[bu] 2
  184. \f[B]fuse_msg_size=INT\f[]: Set the max number of pages per FUSE
  185. message.
  186. Only available on Linux >= 4.20 and ignored otherwise.
  187. (min: 1; max: 256; default: 256)
  188. .IP \[bu] 2
  189. \f[B]threads=INT\f[]: Number of threads to use in multithreaded mode.
  190. When set to zero it will attempt to discover and use the number of
  191. logical cores.
  192. If the lookup fails it will fall back to using 4.
  193. If the thread count is set negative it will look up the number of cores
  194. then divide by the absolute value.
  195. ie.
  196. threads=\-2 on an 8 core machine will result in 8 / 2 = 4 threads.
  197. There will always be at least 1 thread.
  198. NOTE: higher number of threads increases parallelism but usually
  199. decreases throughput.
  200. (default: 0)
  201. .IP \[bu] 2
  202. \f[B]fsname=STR\f[]: Sets the name of the filesystem as seen in
  203. \f[B]mount\f[], \f[B]df\f[], etc.
  204. Defaults to a list of the source paths concatenated together with the
  205. longest common prefix removed.
  206. .IP \[bu] 2
  207. \f[B]func.FUNC=POLICY\f[]: Sets the specific FUSE function\[aq]s policy.
  208. See below for the list of value types.
  209. Example: \f[B]func.getattr=newest\f[]
  210. .IP \[bu] 2
  211. \f[B]category.CATEGORY=POLICY\f[]: Sets policy of all FUSE functions in
  212. the provided category.
  213. See POLICIES section for defaults.
  214. Example: \f[B]category.create=mfs\f[]
  215. .IP \[bu] 2
  216. \f[B]cache.open=INT\f[]: \[aq]open\[aq] policy cache timeout in seconds.
  217. (default: 0)
  218. .IP \[bu] 2
  219. \f[B]cache.statfs=INT\f[]: \[aq]statfs\[aq] cache timeout in seconds.
  220. (default: 0)
  221. .IP \[bu] 2
  222. \f[B]cache.attr=INT\f[]: File attribute cache timeout in seconds.
  223. (default: 1)
  224. .IP \[bu] 2
  225. \f[B]cache.entry=INT\f[]: File name lookup cache timeout in seconds.
  226. (default: 1)
  227. .IP \[bu] 2
  228. \f[B]cache.negative_entry=INT\f[]: Negative file name lookup cache
  229. timeout in seconds.
  230. (default: 0)
  231. .IP \[bu] 2
  232. \f[B]cache.files=libfuse|off|partial|full|auto\-full\f[]: File page
  233. caching mode (default: libfuse)
  234. .IP \[bu] 2
  235. \f[B]cache.symlinks=BOOL\f[]: Cache symlinks (if supported by kernel)
  236. (default: false)
  237. .IP \[bu] 2
  238. \f[B]cache.readdir=BOOL\f[]: Cache readdir (if supported by kernel)
  239. (default: false)
  240. .IP \[bu] 2
  241. \f[B]direct_io\f[]: deprecated \- Bypass page cache.
  242. Use \f[C]cache.files=off\f[] instead.
  243. (default: false)
  244. .IP \[bu] 2
  245. \f[B]kernel_cache\f[]: deprecated \- Do not invalidate data cache on
  246. file open.
  247. Use \f[C]cache.files=full\f[] instead.
  248. (default: false)
  249. .IP \[bu] 2
  250. \f[B]auto_cache\f[]: deprecated \- Invalidate data cache if file mtime
  251. or size change.
  252. Use \f[C]cache.files=auto\-full\f[] instead.
  253. (default: false)
  254. .IP \[bu] 2
  255. \f[B]async_read\f[]: deprecated \- Perform reads asynchronously.
  256. Use \f[C]async_read=true\f[] instead.
  257. .IP \[bu] 2
  258. \f[B]sync_read\f[]: deprecated \- Perform reads synchronously.
  259. Use \f[C]async_read=false\f[] instead.
  260. .PP
  261. \f[B]NOTE:\f[] Options are evaluated in the order listed so if the
  262. options are \f[B]func.rmdir=rand,category.action=ff\f[] the
  263. \f[B]action\f[] category setting will override the \f[B]rmdir\f[]
  264. setting.
  265. .SS Value Types
  266. .IP \[bu] 2
  267. BOOL = \[aq]true\[aq] | \[aq]false\[aq]
  268. .IP \[bu] 2
  269. INT = [0,MAX_INT]
  270. .IP \[bu] 2
  271. SIZE = \[aq]NNM\[aq]; NN = INT, M = \[aq]K\[aq] | \[aq]M\[aq] |
  272. \[aq]G\[aq] | \[aq]T\[aq]
  273. .IP \[bu] 2
  274. STR = string
  275. .IP \[bu] 2
  276. FUNC = FUSE function
  277. .IP \[bu] 2
  278. CATEGORY = FUSE function category
  279. .IP \[bu] 2
  280. POLICY = mergerfs function policy
  281. .SS branches
  282. .PP
  283. The \[aq]branches\[aq] (formerly \[aq]srcmounts\[aq]) argument is a
  284. colon (\[aq]:\[aq]) delimited list of paths to be pooled together.
  285. It does not matter if the paths are on the same or different drives nor
  286. does it matter the filesystem.
  287. Used and available space will not be duplicated for paths on the same
  288. device and any features which aren\[aq]t supported by the underlying
  289. filesystem (such as file attributes or extended attributes) will return
  290. the appropriate errors.
  291. .PP
  292. To make it easier to include multiple branches mergerfs supports
  293. globbing (http://linux.die.net/man/7/glob).
  294. \f[B]The globbing tokens MUST be escaped when using via the shell else
  295. the shell itself will apply the glob itself.\f[]
  296. .PP
  297. Each branch can have a suffix of \f[C]=RW\f[] (read / write),
  298. \f[C]=RO\f[] (read\-only), or \f[C]=NC\f[] (no create).
  299. These suffixes work with globs as well and will apply to each path
  300. found.
  301. \f[C]RW\f[] is the default behavior and those paths will be eligible for
  302. all policy categories.
  303. \f[C]RO\f[] will exclude those paths from \f[C]create\f[] and
  304. \f[C]action\f[] policies (just as a filesystem being mounted \f[C]ro\f[]
  305. would).
  306. \f[C]NC\f[] will exclude those paths from \f[C]create\f[] policies (you
  307. can\[aq]t create but you can change / delete).
  308. .IP
  309. .nf
  310. \f[C]
  311. #\ mergerfs\ \-o\ allow_other,use_ino\ /mnt/disk\\*:/mnt/cdrom\ /media/drives
  312. \f[]
  313. .fi
  314. .PP
  315. The above line will use all mount points in /mnt prefixed with
  316. \f[B]disk\f[] and the \f[B]cdrom\f[].
  317. .PP
  318. To have the pool mounted at boot or otherwise accessable from related
  319. tools use \f[B]/etc/fstab\f[].
  320. .IP
  321. .nf
  322. \f[C]
  323. #\ <file\ system>\ \ \ \ \ \ \ \ <mount\ point>\ \ <type>\ \ \ \ \ \ \ \ \ <options>\ \ \ \ \ \ \ \ \ \ \ \ \ <dump>\ \ <pass>
  324. /mnt/disk*:/mnt/cdrom\ \ /media/drives\ \ fuse.mergerfs\ \ allow_other,use_ino\ \ \ 0\ \ \ \ \ \ \ 0
  325. \f[]
  326. .fi
  327. .PP
  328. \f[B]NOTE:\f[] the globbing is done at mount or xattr update time (see
  329. below).
  330. If a new directory is added matching the glob after the fact it will not
  331. be automatically included.
  332. .PP
  333. \f[B]NOTE:\f[] for mounting via \f[B]fstab\f[] to work you must have
  334. \f[B]mount.fuse\f[] installed.
  335. For Ubuntu/Debian it is included in the \f[B]fuse\f[] package.
  336. .SS fuse_msg_size
  337. .PP
  338. FUSE applications communicate with the kernel over a special character
  339. device: \f[C]/dev/fuse\f[].
  340. A large portion of the overhead associated with FUSE is the cost of
  341. going back and forth from user space and kernel space over that device.
  342. Generally speaking the fewer trips needed the better the performance
  343. will be.
  344. Reducing the number of trips can be done a number of ways.
  345. Kernel level caching and increasing message sizes being two significant
  346. ones.
  347. When it comes to reads and writes if the message size is doubled the
  348. number of trips are appoximately halved.
  349. .PP
  350. In Linux 4.20 a new feature was added allowing the negotiation of the
  351. max message size.
  352. Since the size is in multiples of
  353. pages (https://en.wikipedia.org/wiki/Page_(computer_memory)) the feature
  354. is called \f[C]max_pages\f[].
  355. There is a maximum \f[C]max_pages\f[] value of 256 (1MiB) and minimum of
  356. 1 (4KiB).
  357. The default used by Linux >=4.20, and hardcoded value used before 4.20,
  358. is 32 (128KiB).
  359. In mergerfs its referred to as \f[C]fuse_msg_size\f[] to make it clear
  360. what it impacts and provide some abstraction.
  361. .PP
  362. Since there should be no downsides to increasing \f[C]fuse_msg_size\f[]
  363. / \f[C]max_pages\f[], outside a minor bump in RAM usage due to larger
  364. message buffers, mergerfs defaults the value to 256.
  365. On kernels before 4.20 the value has no effect.
  366. The reason the value is configurable is to enable experimentation and
  367. benchmarking.
  368. See the \f[C]nullrw\f[] section for benchmarking examples.
  369. .SS symlinkify
  370. .PP
  371. Due to the levels of indirection introduced by mergerfs and the
  372. underlying technology FUSE there can be varying levels of performance
  373. degredation.
  374. This feature will turn non\-directories which are not writable into
  375. symlinks to the original file found by the \f[C]readlink\f[] policy
  376. after the mtime and ctime are older than the timeout.
  377. .PP
  378. \f[B]WARNING:\f[] The current implementation has a known issue in which
  379. if the file is open and being used when the file is converted to a
  380. symlink then the application which has that file open will receive an
  381. error when using it.
  382. This is unlikely to occur in practice but is something to keep in mind.
  383. .PP
  384. \f[B]WARNING:\f[] Some backup solutions, such as CrashPlan, do not
  385. backup the target of a symlink.
  386. If using this feature it will be necessary to point any backup software
  387. to the original drives or configure the software to follow symlinks if
  388. such an option is available.
  389. Alternatively create two mounts.
  390. One for backup and one for general consumption.
  391. .SS nullrw
  392. .PP
  393. Due to how FUSE works there is an overhead to all requests made to a
  394. FUSE filesystem.
  395. Meaning that even a simple passthrough will have some slowdown.
  396. However, generally the overhead is minimal in comparison to the cost of
  397. the underlying I/O.
  398. By disabling the underlying I/O we can test the theoretical performance
  399. boundries.
  400. .PP
  401. By enabling \f[C]nullrw\f[] mergerfs will work as it always does
  402. \f[B]except\f[] that all reads and writes will be no\-ops.
  403. A write will succeed (the size of the write will be returned as if it
  404. were successful) but mergerfs does nothing with the data it was given.
  405. Similarly a read will return the size requested but won\[aq]t touch the
  406. buffer.
  407. .PP
  408. Example:
  409. .IP
  410. .nf
  411. \f[C]
  412. $\ dd\ if=/dev/zero\ of=/path/to/mergerfs/mount/benchmark\ ibs=1M\ obs=512\ count=1024\ iflag=dsync,nocache\ oflag=dsync,nocache\ conv=fdatasync\ status=progress
  413. 1024+0\ records\ in
  414. 2097152+0\ records\ out
  415. 1073741824\ bytes\ (1.1\ GB,\ 1.0\ GiB)\ copied,\ 15.4067\ s,\ 69.7\ MB/s
  416. $\ dd\ if=/dev/zero\ of=/path/to/mergerfs/mount/benchmark\ ibs=1M\ obs=1M\ count=1024\ iflag=dsync,nocache\ oflag=dsync,nocache\ conv=fdatasync\ status=progress
  417. 1024+0\ records\ in
  418. 1024+0\ records\ out
  419. 1073741824\ bytes\ (1.1\ GB,\ 1.0\ GiB)\ copied,\ 0.219585\ s,\ 4.9\ GB/s
  420. $\ dd\ if=/path/to/mergerfs/mount/benchmark\ of=/dev/null\ bs=512\ count=102400\ iflag=dsync,nocache\ oflag=dsync,nocache\ conv=fdatasync\ status=progress
  421. 102400+0\ records\ in
  422. 102400+0\ records\ out
  423. 52428800\ bytes\ (52\ MB,\ 50\ MiB)\ copied,\ 0.757991\ s,\ 69.2\ MB/s
  424. $\ dd\ if=/path/to/mergerfs/mount/benchmark\ of=/dev/null\ bs=1M\ count=1024\ iflag=dsync,nocache\ oflag=dsync,nocache\ conv=fdatasync\ status=progress
  425. 1024+0\ records\ in
  426. 1024+0\ records\ out
  427. 1073741824\ bytes\ (1.1\ GB,\ 1.0\ GiB)\ copied,\ 0.18405\ s,\ 5.8\ GB/s
  428. \f[]
  429. .fi
  430. .PP
  431. It\[aq]s important to test with different \f[C]obs\f[] (output block
  432. size) values since the relative overhead is greater with smaller values.
  433. As you can see above the size of a read or write can massively impact
  434. theoretical performance.
  435. If an application performs much worse through mergerfs it could very
  436. well be that it doesn\[aq]t optimally size its read and write requests.
  437. In such cases contact the mergerfs author so it can be investigated.
  438. .SS xattr
  439. .PP
  440. Runtime extended attribute support can be managed via the \f[C]xattr\f[]
  441. option.
  442. By default it will passthrough any xattr calls.
  443. Given xattr support is rarely used and can have significant performance
  444. implications mergerfs allows it to be disabled at runtime.
  445. .PP
  446. \f[C]noattr\f[] will cause mergerfs to short circuit all xattr calls and
  447. return ENOATTR where appropriate.
  448. mergerfs still gets all the requests but they will not be forwarded on
  449. to the underlying filesystems.
  450. The runtime control will still function in this mode.
  451. .PP
  452. \f[C]nosys\f[] will cause mergerfs to return ENOSYS for any xattr call.
  453. The difference with \f[C]noattr\f[] is that the kernel will cache this
  454. fact and itself short circuit future calls.
  455. This will be more efficient than \f[C]noattr\f[] but will cause
  456. mergerfs\[aq] runtime control via the hidden file to stop working.
  457. .SH FUNCTIONS / POLICIES / CATEGORIES
  458. .PP
  459. The POSIX filesystem API is made up of a number of functions.
  460. \f[B]creat\f[], \f[B]stat\f[], \f[B]chown\f[], etc.
  461. In mergerfs most of the core functions are grouped into 3 categories:
  462. \f[B]action\f[], \f[B]create\f[], and \f[B]search\f[].
  463. These functions and categories can be assigned a policy which dictates
  464. what file or directory is chosen when performing that behavior.
  465. Any policy can be assigned to a function or category though some may not
  466. be very useful in practice.
  467. For instance: \f[B]rand\f[] (random) may be useful for file creation
  468. (create) but could lead to very odd behavior if used for \f[C]chmod\f[]
  469. if there were more than one copy of the file.
  470. .PP
  471. Some functions, listed in the category \f[C]N/A\f[] below, can not be
  472. assigned the normal policies.
  473. All functions which work on file handles use the handle which was
  474. acquired by \f[C]open\f[] or \f[C]create\f[].
  475. \f[C]readdir\f[] has no real need for a policy given the purpose is
  476. merely to return a list of entries in a directory.
  477. \f[C]statfs\f[]\[aq]s behavior can be modified via other options.
  478. That said many times the current FUSE kernel driver will not always
  479. provide the file handle when a client calls \f[C]fgetattr\f[],
  480. \f[C]fchown\f[], \f[C]fchmod\f[], \f[C]futimens\f[], \f[C]ftruncate\f[],
  481. etc.
  482. This means it will call the regular, path based, versions.
  483. .PP
  484. When using policies which are based on a branch\[aq]s available space
  485. the base path provided is used.
  486. Not the full path to the file in question.
  487. Meaning that sub mounts won\[aq]t be considered in the space
  488. calculations.
  489. The reason is that it doesn\[aq]t really work for non\-path preserving
  490. policies and can lead to non\-obvious behaviors.
  491. .SS Function / Category classifications
  492. .PP
  493. .TS
  494. tab(@);
  495. lw(7.9n) lw(62.1n).
  496. T{
  497. Category
  498. T}@T{
  499. FUSE Functions
  500. T}
  501. _
  502. T{
  503. action
  504. T}@T{
  505. chmod, chown, link, removexattr, rename, rmdir, setxattr, truncate,
  506. unlink, utimens
  507. T}
  508. T{
  509. create
  510. T}@T{
  511. create, mkdir, mknod, symlink
  512. T}
  513. T{
  514. search
  515. T}@T{
  516. access, getattr, getxattr, ioctl (directories), listxattr, open,
  517. readlink
  518. T}
  519. T{
  520. N/A
  521. T}@T{
  522. fchmod, fchown, futimens, ftruncate, fallocate, fgetattr, fsync, ioctl
  523. (files), read, readdir, release, statfs, write, copy_file_range
  524. T}
  525. .TE
  526. .PP
  527. In cases where something may be searched (to confirm a directory exists
  528. across all source mounts) \f[B]getattr\f[] will be used.
  529. .SS Path Preservation
  530. .PP
  531. Policies, as described below, are of two basic types.
  532. \f[C]path\ preserving\f[] and \f[C]non\-path\ preserving\f[].
  533. .PP
  534. All policies which start with \f[C]ep\f[] (\f[B]epff\f[],
  535. \f[B]eplfs\f[], \f[B]eplus\f[], \f[B]epmfs\f[], \f[B]eprand\f[]) are
  536. \f[C]path\ preserving\f[].
  537. \f[C]ep\f[] stands for \f[C]existing\ path\f[].
  538. .PP
  539. A path preserving policy will only consider drives where the relative
  540. path being accessed already exists.
  541. .PP
  542. When using non\-path preserving policies paths will be cloned to target
  543. drives as necessary.
  544. .SS Filters
  545. .PP
  546. Policies basically search branches and create a list of files / paths
  547. for functions to work on.
  548. The policy is responsible for filtering and sorting.
  549. The policy type defines the sorting but filtering is mostly uniform as
  550. described below.
  551. .IP \[bu] 2
  552. No \f[B]search\f[] policies filter.
  553. .IP \[bu] 2
  554. All \f[B]action\f[] policies will filter out branches which are mounted
  555. \f[B]read\-only\f[] or tagged as \f[B]RO (read\-only)\f[].
  556. .IP \[bu] 2
  557. All \f[B]create\f[] policies will filter out branches which are mounted
  558. \f[B]read\-only\f[], tagged \f[B]RO (read\-only)\f[] or \f[B]NC (no
  559. create)\f[], or has available space less than \f[C]minfreespace\f[].
  560. .PP
  561. If all branches are filtered an error will be returned.
  562. Typically \f[B]EROFS\f[] or \f[B]ENOSPC\f[] depending on the reasons.
  563. .SS Policy descriptions
  564. .PP
  565. .TS
  566. tab(@);
  567. lw(16.6n) lw(53.4n).
  568. T{
  569. Policy
  570. T}@T{
  571. Description
  572. T}
  573. _
  574. T{
  575. all
  576. T}@T{
  577. Search category: same as \f[B]epall\f[].
  578. Action category: same as \f[B]epall\f[].
  579. Create category: for \f[B]mkdir\f[], \f[B]mknod\f[], and
  580. \f[B]symlink\f[] it will apply to all branches.
  581. \f[B]create\f[] works like \f[B]ff\f[].
  582. T}
  583. T{
  584. epall (existing path, all)
  585. T}@T{
  586. Search category: same as \f[B]epff\f[] (but more expensive because it
  587. doesn\[aq]t stop after finding a valid branch).
  588. Action category: apply to all found.
  589. Create category: for \f[B]mkdir\f[], \f[B]mknod\f[], and
  590. \f[B]symlink\f[] it will apply to all found.
  591. \f[B]create\f[] works like \f[B]epff\f[] (but more expensive because it
  592. doesn\[aq]t stop after finding a valid branch).
  593. T}
  594. T{
  595. epff (existing path, first found)
  596. T}@T{
  597. Given the order of the branches, as defined at mount time or configured
  598. at runtime, act on the first one found where the relative path exists.
  599. T}
  600. T{
  601. eplfs (existing path, least free space)
  602. T}@T{
  603. Of all the branches on which the relative path exists choose the drive
  604. with the least free space.
  605. T}
  606. T{
  607. eplus (existing path, least used space)
  608. T}@T{
  609. Of all the branches on which the relative path exists choose the drive
  610. with the least used space.
  611. T}
  612. T{
  613. epmfs (existing path, most free space)
  614. T}@T{
  615. Of all the branches on which the relative path exists choose the drive
  616. with the most free space.
  617. T}
  618. T{
  619. eprand (existing path, random)
  620. T}@T{
  621. Calls \f[B]epall\f[] and then randomizes.
  622. Returns 1.
  623. T}
  624. T{
  625. erofs
  626. T}@T{
  627. Exclusively return \f[B]\-1\f[] with \f[B]errno\f[] set to
  628. \f[B]EROFS\f[] (read\-only filesystem).
  629. T}
  630. T{
  631. ff (first found)
  632. T}@T{
  633. Search category: same as \f[B]epff\f[].
  634. Action category: same as \f[B]epff\f[].
  635. Create category: Given the order of the drives, as defined at mount time
  636. or configured at runtime, act on the first one found.
  637. T}
  638. T{
  639. lfs (least free space)
  640. T}@T{
  641. Search category: same as \f[B]eplfs\f[].
  642. Action category: same as \f[B]eplfs\f[].
  643. Create category: Pick the drive with the least available free space.
  644. T}
  645. T{
  646. lus (least used space)
  647. T}@T{
  648. Search category: same as \f[B]eplus\f[].
  649. Action category: same as \f[B]eplus\f[].
  650. Create category: Pick the drive with the least used space.
  651. T}
  652. T{
  653. mfs (most free space)
  654. T}@T{
  655. Search category: same as \f[B]epmfs\f[].
  656. Action category: same as \f[B]epmfs\f[].
  657. Create category: Pick the drive with the most available free space.
  658. T}
  659. T{
  660. newest
  661. T}@T{
  662. Pick the file / directory with the largest mtime.
  663. T}
  664. T{
  665. rand (random)
  666. T}@T{
  667. Calls \f[B]all\f[] and then randomizes.
  668. Returns 1.
  669. T}
  670. .TE
  671. .SS Defaults
  672. .PP
  673. .TS
  674. tab(@);
  675. l l.
  676. T{
  677. Category
  678. T}@T{
  679. Policy
  680. T}
  681. _
  682. T{
  683. action
  684. T}@T{
  685. epall
  686. T}
  687. T{
  688. create
  689. T}@T{
  690. epmfs
  691. T}
  692. T{
  693. search
  694. T}@T{
  695. ff
  696. T}
  697. .TE
  698. .SS ioctl
  699. .PP
  700. When \f[C]ioctl\f[] is used with an open file then it will use the file
  701. handle which was created at the original \f[C]open\f[] call.
  702. However, when using \f[C]ioctl\f[] with a directory mergerfs will use
  703. the \f[C]open\f[] policy to find the directory to act on.
  704. .SS unlink
  705. .PP
  706. In FUSE there is an opaque "file handle" which is created by
  707. \f[C]open\f[], \f[C]create\f[], or \f[C]opendir\f[], passed to the
  708. kernel, and then is passed back to the FUSE userland application by the
  709. kernel.
  710. Unfortunately, the FUSE kernel driver does not always send the file
  711. handle when it theoretically could/should.
  712. This complicates certain behaviors / workflows particularly in the high
  713. level API.
  714. As a result mergerfs is currently doing a few hacky things.
  715. .PP
  716. libfuse2 and libfuse3, when using the high level API, will rename names
  717. to \f[C]\&.fuse_hiddenXXXXXX\f[] if the file is open when unlinked or
  718. renamed over.
  719. It does this so the file is still available when a request referencing
  720. the now missing file is made.
  721. This file however keeps a \f[C]rmdir\f[] from succeeding and can be
  722. picked up by software reading directories.
  723. .PP
  724. The change mergerfs has done is that if a file is open when an unlink or
  725. rename happens it will open the file and keep it open till closed by all
  726. those who opened it prior.
  727. When a request comes in referencing that file and it doesn\[aq]t include
  728. a file handle it will instead use the file handle created at
  729. unlink/rename time.
  730. .PP
  731. This won\[aq]t result in technically proper behavior but close enough
  732. for many usecases.
  733. .PP
  734. The plan is to rewrite mergerfs to use the low level API so these
  735. invasive libfuse changes are no longer necessary.
  736. .SS rename & link
  737. .PP
  738. \f[B]NOTE:\f[] If you\[aq]re receiving errors from software when files
  739. are moved / renamed / linked then you should consider changing the
  740. create policy to one which is \f[B]not\f[] path preserving, enabling
  741. \f[C]ignorepponrename\f[], or contacting the author of the offending
  742. software and requesting that \f[C]EXDEV\f[] be properly handled.
  743. .PP
  744. \f[C]rename\f[] and \f[C]link\f[] are tricky functions in a union
  745. filesystem.
  746. \f[C]rename\f[] only works within a single filesystem or device.
  747. If a rename can\[aq]t be done atomically due to the source and
  748. destination paths existing on different mount points it will return
  749. \f[B]\-1\f[] with \f[B]errno = EXDEV\f[] (cross device).
  750. So if a \f[C]rename\f[]\[aq]s source and target are on different drives
  751. within the pool it creates an issue.
  752. .PP
  753. Originally mergerfs would return EXDEV whenever a rename was requested
  754. which was cross directory in any way.
  755. This made the code simple and was technically complient with POSIX
  756. requirements.
  757. However, many applications fail to handle EXDEV at all and treat it as a
  758. normal error or otherwise handle it poorly.
  759. Such apps include: gvfsd\-fuse v1.20.3 and prior, Finder / CIFS/SMB
  760. client in Apple OSX 10.9+, NZBGet, Samba\[aq]s recycling bin feature.
  761. .PP
  762. As a result a compromise was made in order to get most software to work
  763. while still obeying mergerfs\[aq] policies.
  764. Below is the basic logic.
  765. .IP \[bu] 2
  766. If using a \f[B]create\f[] policy which tries to preserve directory
  767. paths (epff,eplfs,eplus,epmfs)
  768. .IP \[bu] 2
  769. Using the \f[B]rename\f[] policy get the list of files to rename
  770. .IP \[bu] 2
  771. For each file attempt rename:
  772. .RS 2
  773. .IP \[bu] 2
  774. If failure with ENOENT run \f[B]create\f[] policy
  775. .IP \[bu] 2
  776. If create policy returns the same drive as currently evaluating then
  777. clone the path
  778. .IP \[bu] 2
  779. Re\-attempt rename
  780. .RE
  781. .IP \[bu] 2
  782. If \f[B]any\f[] of the renames succeed the higher level rename is
  783. considered a success
  784. .IP \[bu] 2
  785. If \f[B]no\f[] renames succeed the first error encountered will be
  786. returned
  787. .IP \[bu] 2
  788. On success:
  789. .RS 2
  790. .IP \[bu] 2
  791. Remove the target from all drives with no source file
  792. .IP \[bu] 2
  793. Remove the source from all drives which failed to rename
  794. .RE
  795. .IP \[bu] 2
  796. If using a \f[B]create\f[] policy which does \f[B]not\f[] try to
  797. preserve directory paths
  798. .IP \[bu] 2
  799. Using the \f[B]rename\f[] policy get the list of files to rename
  800. .IP \[bu] 2
  801. Using the \f[B]getattr\f[] policy get the target path
  802. .IP \[bu] 2
  803. For each file attempt rename:
  804. .RS 2
  805. .IP \[bu] 2
  806. If the source drive != target drive:
  807. .IP \[bu] 2
  808. Clone target path from target drive to source drive
  809. .IP \[bu] 2
  810. Rename
  811. .RE
  812. .IP \[bu] 2
  813. If \f[B]any\f[] of the renames succeed the higher level rename is
  814. considered a success
  815. .IP \[bu] 2
  816. If \f[B]no\f[] renames succeed the first error encountered will be
  817. returned
  818. .IP \[bu] 2
  819. On success:
  820. .RS 2
  821. .IP \[bu] 2
  822. Remove the target from all drives with no source file
  823. .IP \[bu] 2
  824. Remove the source from all drives which failed to rename
  825. .RE
  826. .PP
  827. The the removals are subject to normal entitlement checks.
  828. .PP
  829. The above behavior will help minimize the likelihood of EXDEV being
  830. returned but it will still be possible.
  831. .PP
  832. \f[B]link\f[] uses the same strategy but without the removals.
  833. .SS readdir
  834. .PP
  835. readdir (http://linux.die.net/man/3/readdir) is different from all other
  836. filesystem functions.
  837. While it could have its own set of policies to tweak its behavior at
  838. this time it provides a simple union of files and directories found.
  839. Remember that any action or information queried about these files and
  840. directories come from the respective function.
  841. For instance: an \f[B]ls\f[] is a \f[B]readdir\f[] and for each
  842. file/directory returned \f[B]getattr\f[] is called.
  843. Meaning the policy of \f[B]getattr\f[] is responsible for choosing the
  844. file/directory which is the source of the metadata you see in an
  845. \f[B]ls\f[].
  846. .SS statfs / statvfs
  847. .PP
  848. statvfs (http://linux.die.net/man/2/statvfs) normalizes the source
  849. drives based on the fragment size and sums the number of adjusted blocks
  850. and inodes.
  851. This means you will see the combined space of all sources.
  852. Total, used, and free.
  853. The sources however are dedupped based on the drive so multiple sources
  854. on the same drive will not result in double counting its space.
  855. Filesystems mounted further down the tree of the branch will not be
  856. included when checking the mount\[aq]s stats.
  857. .PP
  858. The options \f[C]statfs\f[] and \f[C]statfs_ignore\f[] can be used to
  859. modify \f[C]statfs\f[] behavior.
  860. .SH BUILDING
  861. .PP
  862. \f[B]NOTE:\f[] Prebuilt packages can be found at:
  863. https://github.com/trapexit/mergerfs/releases
  864. .PP
  865. First get the code from github (https://github.com/trapexit/mergerfs).
  866. .IP
  867. .nf
  868. \f[C]
  869. $\ git\ clone\ https://github.com/trapexit/mergerfs.git
  870. $\ #\ or
  871. $\ wget\ https://github.com/trapexit/mergerfs/releases/download/<ver>/mergerfs\-<ver>.tar.gz
  872. \f[]
  873. .fi
  874. .SS Debian / Ubuntu
  875. .IP
  876. .nf
  877. \f[C]
  878. $\ cd\ mergerfs
  879. $\ sudo\ tools/install\-build\-pkgs
  880. $\ make\ deb
  881. $\ sudo\ dpkg\ \-i\ ../mergerfs_version_arch.deb
  882. \f[]
  883. .fi
  884. .SS Fedora
  885. .IP
  886. .nf
  887. \f[C]
  888. $\ su\ \-
  889. #\ cd\ mergerfs
  890. #\ tools/install\-build\-pkgs
  891. #\ make\ rpm
  892. #\ rpm\ \-i\ rpmbuild/RPMS/<arch>/mergerfs\-<verion>.<arch>.rpm
  893. \f[]
  894. .fi
  895. .SS Generically
  896. .PP
  897. Have git, g++, make, python installed.
  898. .IP
  899. .nf
  900. \f[C]
  901. $\ cd\ mergerfs
  902. $\ make
  903. $\ sudo\ make\ install
  904. \f[]
  905. .fi
  906. .SS Build options
  907. .IP
  908. .nf
  909. \f[C]
  910. $\ make\ help
  911. usage:\ make
  912. make\ USE_XATTR=0\ \ \ \ \ \ \-\ build\ program\ without\ xattrs\ functionality
  913. make\ STATIC=1\ \ \ \ \ \ \ \ \ \-\ build\ static\ binary
  914. make\ LTO=1\ \ \ \ \ \ \ \ \ \ \ \ \-\ build\ with\ link\ time\ optimization
  915. \f[]
  916. .fi
  917. .SH RUNTIME CONFIG
  918. .SS .mergerfs pseudo file
  919. .IP
  920. .nf
  921. \f[C]
  922. <mountpoint>/.mergerfs
  923. \f[]
  924. .fi
  925. .PP
  926. There is a pseudo file available at the mount point which allows for the
  927. runtime modification of certain \f[B]mergerfs\f[] options.
  928. The file will not show up in \f[B]readdir\f[] but can be
  929. \f[B]stat\f[]\[aq]ed and manipulated via
  930. {list,get,set}xattrs (http://linux.die.net/man/2/listxattr) calls.
  931. .PP
  932. Any changes made at runtime are \f[B]not\f[] persisted.
  933. If you wish for values to persist they must be included as options
  934. wherever you configure the mounting of mergerfs (/etc/fstab).
  935. .SS Keys
  936. .PP
  937. Use \f[C]xattr\ \-l\ /mountpoint/.mergerfs\f[] to see all supported
  938. keys.
  939. Some are informational and therefore read\-only.
  940. \f[C]setxattr\f[] will return EINVAL on read\-only keys.
  941. .SS Values
  942. .PP
  943. Same as the command line.
  944. .SS user.mergerfs.branches
  945. .PP
  946. \f[B]NOTE:\f[] formerly \f[C]user.mergerfs.srcmounts\f[] but said key is
  947. still supported.
  948. .PP
  949. Used to query or modify the list of branches.
  950. When modifying there are several shortcuts to easy manipulation of the
  951. list.
  952. .PP
  953. .TS
  954. tab(@);
  955. l l.
  956. T{
  957. Value
  958. T}@T{
  959. Description
  960. T}
  961. _
  962. T{
  963. [list]
  964. T}@T{
  965. set
  966. T}
  967. T{
  968. +<[list]
  969. T}@T{
  970. prepend
  971. T}
  972. T{
  973. +>[list]
  974. T}@T{
  975. append
  976. T}
  977. T{
  978. \-[list]
  979. T}@T{
  980. remove all values provided
  981. T}
  982. T{
  983. \-<
  984. T}@T{
  985. remove first in list
  986. T}
  987. T{
  988. \->
  989. T}@T{
  990. remove last in list
  991. T}
  992. .TE
  993. .PP
  994. \f[C]xattr\ \-w\ user.mergerfs.branches\ +</mnt/drive3\ /mnt/pool/.mergerfs\f[]
  995. .PP
  996. The \f[C]=NC\f[], \f[C]=RO\f[], \f[C]=RW\f[] syntax works just as on the
  997. command line.
  998. .SS Example
  999. .IP
  1000. .nf
  1001. \f[C]
  1002. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-l\ .mergerfs
  1003. user.mergerfs.branches:\ /mnt/a=RW:/mnt/b=RW
  1004. user.mergerfs.minfreespace:\ 4294967295
  1005. user.mergerfs.moveonenospc:\ false
  1006. \&...
  1007. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.category.search\ .mergerfs
  1008. ff
  1009. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-w\ user.mergerfs.category.search\ newest\ .mergerfs
  1010. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.category.search\ .mergerfs
  1011. newest
  1012. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-w\ user.mergerfs.branches\ +/mnt/c\ .mergerfs
  1013. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.branches\ .mergerfs
  1014. /mnt/a:/mnt/b:/mnt/c
  1015. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-w\ user.mergerfs.branches\ =/mnt/c\ .mergerfs
  1016. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.branches\ .mergerfs
  1017. /mnt/c
  1018. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-w\ user.mergerfs.branches\ \[aq]+</mnt/a:/mnt/b\[aq]\ .mergerfs
  1019. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.branches\ .mergerfs
  1020. /mnt/a:/mnt/b:/mnt/c
  1021. \f[]
  1022. .fi
  1023. .SS file / directory xattrs
  1024. .PP
  1025. While they won\[aq]t show up when using
  1026. listxattr (http://linux.die.net/man/2/listxattr) \f[B]mergerfs\f[]
  1027. offers a number of special xattrs to query information about the files
  1028. served.
  1029. To access the values you will need to issue a
  1030. getxattr (http://linux.die.net/man/2/getxattr) for one of the following:
  1031. .IP \[bu] 2
  1032. \f[B]user.mergerfs.basepath\f[]: the base mount point for the file given
  1033. the current getattr policy
  1034. .IP \[bu] 2
  1035. \f[B]user.mergerfs.relpath\f[]: the relative path of the file from the
  1036. perspective of the mount point
  1037. .IP \[bu] 2
  1038. \f[B]user.mergerfs.fullpath\f[]: the full path of the original file
  1039. given the getattr policy
  1040. .IP \[bu] 2
  1041. \f[B]user.mergerfs.allpaths\f[]: a NUL (\[aq]\[aq]) separated list of
  1042. full paths to all files found
  1043. .IP
  1044. .nf
  1045. \f[C]
  1046. [trapexit:/mnt/mergerfs]\ $\ ls
  1047. A\ B\ C
  1048. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.fullpath\ A
  1049. /mnt/a/full/path/to/A
  1050. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.basepath\ A
  1051. /mnt/a
  1052. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.relpath\ A
  1053. /full/path/to/A
  1054. [trapexit:/mnt/mergerfs]\ $\ xattr\ \-p\ user.mergerfs.allpaths\ A\ |\ tr\ \[aq]\\0\[aq]\ \[aq]\\n\[aq]
  1055. /mnt/a/full/path/to/A
  1056. /mnt/b/full/path/to/A
  1057. \f[]
  1058. .fi
  1059. .SH UPGRADE
  1060. .PP
  1061. mergerfs can be upgraded live by mounting on top of the previous
  1062. version.
  1063. Simply install the new version of mergerfs and follow the instructions
  1064. below.
  1065. .PP
  1066. Add \f[C]nonempty\f[] to your mergerfs option list and call mergerfs
  1067. again or if using \f[C]/etc/fstab\f[] call for it to mount again.
  1068. Existing open files and such will continue to work fine though they
  1069. won\[aq]t see runtime changes since any such change would be the new
  1070. mount.
  1071. If you plan on changing settings with the new mount you should / could
  1072. apply those before mounting the new version.
  1073. .IP
  1074. .nf
  1075. \f[C]
  1076. $\ sudo\ mount\ /mnt/mergerfs
  1077. $\ mount\ |\ grep\ mergerfs
  1078. media\ on\ /mnt/mergerfs\ type\ fuse.mergerfs\ (rw,relatime,user_id=0,group_id=0,default_permissions,allow_other)
  1079. media\ on\ /mnt/mergerfs\ type\ fuse.mergerfs\ (rw,relatime,user_id=0,group_id=0,default_permissions,allow_other)
  1080. \f[]
  1081. .fi
  1082. .SH TOOLING
  1083. .IP \[bu] 2
  1084. https://github.com/trapexit/mergerfs\-tools
  1085. .IP \[bu] 2
  1086. mergerfs.ctl: A tool to make it easier to query and configure mergerfs
  1087. at runtime
  1088. .IP \[bu] 2
  1089. mergerfs.fsck: Provides permissions and ownership auditing and the
  1090. ability to fix them
  1091. .IP \[bu] 2
  1092. mergerfs.dedup: Will help identify and optionally remove duplicate files
  1093. .IP \[bu] 2
  1094. mergerfs.dup: Ensure there are at least N copies of a file across the
  1095. pool
  1096. .IP \[bu] 2
  1097. mergerfs.balance: Rebalance files across drives by moving them from the
  1098. most filled to the least filled
  1099. .IP \[bu] 2
  1100. mergerfs.consolidate: move files within a single mergerfs directory to
  1101. the drive with most free space
  1102. .IP \[bu] 2
  1103. mergerfs.mktrash: Creates FreeDesktop.org Trash specification compatible
  1104. directories on a mergerfs mount
  1105. .IP \[bu] 2
  1106. https://github.com/trapexit/scorch
  1107. .IP \[bu] 2
  1108. scorch: A tool to help discover silent corruption of files and keep
  1109. track of files
  1110. .IP \[bu] 2
  1111. https://github.com/trapexit/bbf
  1112. .IP \[bu] 2
  1113. bbf (bad block finder): a tool to scan for and \[aq]fix\[aq] hard drive
  1114. bad blocks and find the files using those blocks
  1115. .SH CACHING
  1116. .SS page caching
  1117. .PP
  1118. https://en.wikipedia.org/wiki/Page_cache
  1119. .PP
  1120. tl;dr: * cache.files=off: Disables page caching.
  1121. Underlying files cached, mergerfs files are not.
  1122. * cache.files=partial: Enables page caching.
  1123. Underlying files cached, mergerfs files cached while open.
  1124. * cache.files=full: Enables page caching.
  1125. Underlying files cached, mergerfs files cached across opens.
  1126. * cache.files=auto\-full: Enables page caching.
  1127. Underlying files cached, mergerfs files cached across opens if mtime and
  1128. size are unchanged since previous open.
  1129. * cache.files=libfuse: follow traditional libfuse \f[C]direct_io\f[],
  1130. \f[C]kernel_cache\f[], and \f[C]auto_cache\f[] arguments.
  1131. .PP
  1132. FUSE, which mergerfs uses, offers a number of page caching modes.
  1133. mergerfs tries to simplify their use via the \f[C]cache.files\f[]
  1134. option.
  1135. It can and should replace usage of \f[C]direct_io\f[],
  1136. \f[C]kernel_cache\f[], and \f[C]auto_cache\f[].
  1137. .PP
  1138. Due to mergerfs using FUSE and therefore being a userland process
  1139. proxying existing filesystems the kernel will double cache the content
  1140. being read and written through mergerfs.
  1141. Once from the underlying filesystem and once from mergerfs (it sees them
  1142. as two separate entities).
  1143. Using \f[C]cache.files=off\f[] will keep the double caching from
  1144. happening by disabling caching of mergerfs but this has the side effect
  1145. that \f[I]all\f[] read and write calls will be passed to mergerfs which
  1146. may be slower than enabling caching, you lose shared \f[C]mmap\f[]
  1147. support which can affect apps such as rtorrent, and no read\-ahead will
  1148. take place.
  1149. The kernel will still cache the underlying filesystem data but that only
  1150. helps so much given mergerfs will still process all requests.
  1151. .PP
  1152. If you do enable file page caching,
  1153. \f[C]cache.files=partial|full|auto\-full\f[], you should also enable
  1154. \f[C]dropcacheonclose\f[] which will cause mergerfs to instruct the
  1155. kernel to flush the underlying file\[aq]s page cache when the file is
  1156. closed.
  1157. This behavior is the same as the rsync fadvise / drop cache patch and
  1158. Feh\[aq]s nocache project.
  1159. .PP
  1160. If most files are read once through and closed (like media) it is best
  1161. to enable \f[C]dropcacheonclose\f[] regardless of caching mode in order
  1162. to minimize buffer bloat.
  1163. .PP
  1164. It is difficult to balance memory usage, cache bloat & duplication, and
  1165. performance.
  1166. Ideally mergerfs would be able to disable caching for the files it
  1167. reads/writes but allow page caching for itself.
  1168. That would limit the FUSE overhead.
  1169. However, there isn\[aq]t a good way to achieve this.
  1170. It would need to open all files with O_DIRECT which places limitations
  1171. on the what underlying filesystems would be supported and complicates
  1172. the code.
  1173. .PP
  1174. kernel documenation:
  1175. https://www.kernel.org/doc/Documentation/filesystems/fuse\-io.txt
  1176. .SS entry & attribute caching
  1177. .PP
  1178. Given the relatively high cost of FUSE due to the kernel <\-> userspace
  1179. round trips there are kernel side caches for file entries and
  1180. attributes.
  1181. The entry cache limits the \f[C]lookup\f[] calls to mergerfs which ask
  1182. if a file exists.
  1183. The attribute cache limits the need to make \f[C]getattr\f[] calls to
  1184. mergerfs which provide file attributes (mode, size, type, etc.).
  1185. As with the page cache these should not be used if the underlying
  1186. filesystems are being manipulated at the same time as it could lead to
  1187. odd behavior or data corruption.
  1188. The options for setting these are \f[C]cache.entry\f[] and
  1189. \f[C]cache.negative_entry\f[] for the entry cache and
  1190. \f[C]cache.attr\f[] for the attributes cache.
  1191. \f[C]cache.negative_entry\f[] refers to the timeout for negative
  1192. responses to lookups (non\-existant files).
  1193. .SS policy caching
  1194. .PP
  1195. Policies are run every time a function (with a policy as mentioned
  1196. above) is called.
  1197. These policies can be expensive depending on mergerfs\[aq] setup and
  1198. client usage patterns.
  1199. Generally we wouldn\[aq]t want to cache policy results because it may
  1200. result in stale responses if the underlying drives are used directly.
  1201. .PP
  1202. The \f[C]open\f[] policy cache will cache the result of an \f[C]open\f[]
  1203. policy for a particular input for \f[C]cache.open\f[] seconds or until
  1204. the file is unlinked.
  1205. Each file close (release) will randomly chose to clean up the cache of
  1206. expired entries.
  1207. .PP
  1208. This cache is really only useful in cases where you have a large number
  1209. of branches and \f[C]open\f[] is called on the same files repeatedly
  1210. (like \f[B]Transmission\f[] which opens and closes a file on every
  1211. read/write presumably to keep file handle usage low).
  1212. .SS statfs caching
  1213. .PP
  1214. Of the syscalls used by mergerfs in policies the \f[C]statfs\f[] /
  1215. \f[C]statvfs\f[] call is perhaps the most expensive.
  1216. It\[aq]s used to find out the available space of a drive and whether it
  1217. is mounted read\-only.
  1218. Depending on the setup and usage pattern these queries can be relatively
  1219. costly.
  1220. When \f[C]cache.statfs\f[] is enabled all calls to \f[C]statfs\f[] by a
  1221. policy will be cached for the number of seconds its set to.
  1222. .PP
  1223. Example: If the create policy is \f[C]mfs\f[] and the timeout is 60 then
  1224. for that 60 seconds the same drive will be returned as the target for
  1225. creates because the available space won\[aq]t be updated for that time.
  1226. .SS symlink caching
  1227. .PP
  1228. As of version 4.20 Linux supports symlink caching.
  1229. Significant performance increases can be had in workloads which use a
  1230. lot of symlinks.
  1231. Setting \f[C]cache.symlinks=true\f[] will result in requesting symlink
  1232. caching from the kernel only if supported.
  1233. As a result its safe to enable it on systems prior to 4.20.
  1234. That said it is disabled by default for now.
  1235. You can see if caching is enabled by querying the xattr
  1236. \f[C]user.mergerfs.cache.symlinks\f[] but given it must be requested at
  1237. startup you can not change it at runtime.
  1238. .SS readdir caching
  1239. .PP
  1240. As of version 4.20 Linux supports readdir caching.
  1241. This can have a significant impact on directory traversal.
  1242. Especially when combined with entry (\f[C]cache.entry\f[]) and attribute
  1243. (\f[C]cache.attr\f[]) caching.
  1244. Setting \f[C]cache.readdir=true\f[] will result in requesting readdir
  1245. caching from the kernel on each \f[C]opendir\f[].
  1246. If the kernel doesn\[aq]t support readdir caching setting the option to
  1247. \f[C]true\f[] has no effect.
  1248. This option is configuarable at runtime via xattr
  1249. \f[C]user.mergerfs.cache.readdir\f[].
  1250. .SS writeback caching
  1251. .PP
  1252. writeback caching is a technique for improving write speeds by batching
  1253. writes at a faster device and then bulk writing to the slower device.
  1254. With FUSE the kernel will wait for a number of writes to be made and
  1255. then send it to the filesystem as one request.
  1256. mergerfs currently uses a modified and vendored libfuse 2.9.7 which does
  1257. not support writeback caching.
  1258. Adding said feature should not be difficult but benchmarking needs to be
  1259. done to see if what effect it will have.
  1260. .SS tiered caching
  1261. .PP
  1262. Some storage technologies support what some call "tiered" caching.
  1263. The placing of usually smaller, faster storage as a transparent cache to
  1264. larger, slower storage.
  1265. NVMe, SSD, Optane in front of traditional HDDs for instance.
  1266. .PP
  1267. MergerFS does not natively support any sort of tiered caching.
  1268. Most users have no use for such a feature and its inclusion would
  1269. complicate the code.
  1270. However, there are a few situations where a cache drive could help with
  1271. a typical mergerfs setup.
  1272. .IP "1." 3
  1273. Fast network, slow drives, many readers: You\[aq]ve a 10+Gbps network
  1274. with many readers and your regular drives can\[aq]t keep up.
  1275. .IP "2." 3
  1276. Fast network, slow drives, small\[aq]ish bursty writes: You have a
  1277. 10+Gbps network and wish to transfer amounts of data less than your
  1278. cache drive but wish to do so quickly.
  1279. .PP
  1280. With #1 its arguable if you should be using mergerfs at all.
  1281. RAID would probably be the better solution.
  1282. If you\[aq]re going to use mergerfs there are other tactics that may
  1283. help: spreading the data across drives (see the mergerfs.dup tool) and
  1284. setting \f[C]func.open=rand\f[], using \f[C]symlinkify\f[], or using
  1285. dm\-cache or a similar technology to add tiered cache to the underlying
  1286. device.
  1287. .PP
  1288. With #2 one could use dm\-cache as well but there is another solution
  1289. which requires only mergerfs and a cronjob.
  1290. .IP "1." 3
  1291. Create 2 mergerfs pools.
  1292. One which includes just the slow drives and one which has both the fast
  1293. drives (SSD,NVME,etc.) and slow drives.
  1294. .IP "2." 3
  1295. The \[aq]cache\[aq] pool should have the cache drives listed first.
  1296. .IP "3." 3
  1297. The best \f[C]create\f[] policies to use for the \[aq]cache\[aq] pool
  1298. would probably be \f[C]ff\f[], \f[C]epff\f[], \f[C]lfs\f[], or
  1299. \f[C]eplfs\f[].
  1300. The latter two under the assumption that the cache drive(s) are far
  1301. smaller than the backing drives.
  1302. If using path preserving policies remember that you\[aq]ll need to
  1303. manually create the core directories of those paths you wish to be
  1304. cached.
  1305. Be sure the permissions are in sync.
  1306. Use \f[C]mergerfs.fsck\f[] to check / correct them.
  1307. You could also tag the slow drives as \f[C]=NC\f[] though that\[aq]d
  1308. mean if the cache drives fill you\[aq]d get "out of space" errors.
  1309. .IP "4." 3
  1310. Enable \f[C]moveonenospc\f[] and set \f[C]minfreespace\f[]
  1311. appropriately.
  1312. Perhaps setting \f[C]minfreespace\f[] to the size of the largest cache
  1313. drive.
  1314. .IP "5." 3
  1315. Set your programs to use the cache pool.
  1316. .IP "6." 3
  1317. Save one of the below scripts or create you\[aq]re own.
  1318. .IP "7." 3
  1319. Use \f[C]cron\f[] (as root) to schedule the command at whatever
  1320. frequency is appropriate for your workflow.
  1321. .SS time based expiring
  1322. .PP
  1323. Move files from cache to backing pool based only on the last time the
  1324. file was accessed.
  1325. Replace \f[C]\-atime\f[] with \f[C]\-amin\f[] if you want minutes rather
  1326. than days.
  1327. May want to use the \f[C]fadvise\f[] / \f[C]\-\-drop\-cache\f[] version
  1328. of rsync or run rsync with the tool "nocache".
  1329. .IP
  1330. .nf
  1331. \f[C]
  1332. #!/bin/bash
  1333. if\ [\ $#\ !=\ 3\ ];\ then
  1334. \ \ echo\ "usage:\ $0\ <cache\-drive>\ <backing\-pool>\ <days\-old>"
  1335. \ \ exit\ 1
  1336. fi
  1337. CACHE="${1}"
  1338. BACKING="${2}"
  1339. N=${3}
  1340. find\ "${CACHE}"\ \-type\ f\ \-atime\ +${N}\ \-printf\ \[aq]%P\\n\[aq]\ |\ \\
  1341. \ \ rsync\ \-\-files\-from=\-\ \-axqHAXWES\ \-\-preallocate\ \-\-remove\-source\-files\ "${CACHE}/"\ "${BACKING}/"
  1342. \f[]
  1343. .fi
  1344. .SS percentage full expiring
  1345. .PP
  1346. Move the oldest file from the cache to the backing pool.
  1347. Continue till below percentage threshold.
  1348. .IP
  1349. .nf
  1350. \f[C]
  1351. #!/bin/bash
  1352. if\ [\ $#\ !=\ 3\ ];\ then
  1353. \ \ echo\ "usage:\ $0\ <cache\-drive>\ <backing\-pool>\ <percentage>"
  1354. \ \ exit\ 1
  1355. fi
  1356. CACHE="${1}"
  1357. BACKING="${2}"
  1358. PERCENTAGE=${3}
  1359. set\ \-o\ errexit
  1360. while\ [\ $(df\ \-\-output=pcent\ "${CACHE}"\ |\ grep\ \-v\ Use\ |\ cut\ \-d\[aq]%\[aq]\ \-f1)\ \-gt\ ${PERCENTAGE}\ ]
  1361. do
  1362. \ \ \ \ FILE=$(find\ "${CACHE}"\ \-type\ f\ \-printf\ \[aq]%A\@\ %P\\n\[aq]\ |\ \\
  1363. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ sort\ |\ \\
  1364. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ head\ \-n\ 1\ |\ \\
  1365. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ cut\ \-d\[aq]\ \[aq]\ \-f2\-)
  1366. \ \ \ \ test\ \-n\ "${FILE}"
  1367. \ \ \ \ rsync\ \-axqHAXWES\ \-\-preallocate\ \-\-remove\-source\-files\ "${CACHE}/./${FILE}"\ "${BACKING}/"
  1368. done
  1369. \f[]
  1370. .fi
  1371. .SH TIPS / NOTES
  1372. .IP \[bu] 2
  1373. \f[B]use_ino\f[] will only work when used with mergerfs 2.18.0 and
  1374. above.
  1375. .IP \[bu] 2
  1376. Run mergerfs as \f[C]root\f[] (with \f[B]allow_other\f[]) unless
  1377. you\[aq]re merging paths which are owned by the same user otherwise
  1378. strange permission issues may arise.
  1379. .IP \[bu] 2
  1380. https://github.com/trapexit/backup\-and\-recovery\-howtos : A set of
  1381. guides / howtos on creating a data storage system, backing it up,
  1382. maintaining it, and recovering from failure.
  1383. .IP \[bu] 2
  1384. If you don\[aq]t see some directories and files you expect in a merged
  1385. point or policies seem to skip drives be sure the user has permission to
  1386. all the underlying directories.
  1387. Use \f[C]mergerfs.fsck\f[] to audit the drive for out of sync
  1388. permissions.
  1389. .IP \[bu] 2
  1390. Do \f[B]not\f[] use \f[C]cache.files=off\f[] or \f[C]direct_io\f[] if
  1391. you expect applications (such as rtorrent) to
  1392. mmap (http://linux.die.net/man/2/mmap) files.
  1393. Shared mmap is not currently supported in FUSE w/ \f[C]direct_io\f[]
  1394. enabled.
  1395. Enabling \f[C]dropcacheonclose\f[] is recommended when
  1396. \f[C]cache.files=partial|full|auto\-full\f[] or
  1397. \f[C]direct_io=false\f[].
  1398. .IP \[bu] 2
  1399. Since POSIX functions give only a singular error or success its
  1400. difficult to determine the proper behavior when applying the function to
  1401. multiple targets.
  1402. \f[B]mergerfs\f[] will return an error only if all attempts of an action
  1403. fail.
  1404. Any success will lead to a success returned.
  1405. This means however that some odd situations may arise.
  1406. .IP \[bu] 2
  1407. Kodi (http://kodi.tv), Plex (http://plex.tv),
  1408. Subsonic (http://subsonic.org), etc.
  1409. can use directory mtime (http://linux.die.net/man/2/stat) to more
  1410. efficiently determine whether to scan for new content rather than simply
  1411. performing a full scan.
  1412. If using the default \f[B]getattr\f[] policy of \f[B]ff\f[] its possible
  1413. those programs will miss an update on account of it returning the first
  1414. directory found\[aq]s \f[B]stat\f[] info and its a later directory on
  1415. another mount which had the \f[B]mtime\f[] recently updated.
  1416. To fix this you will want to set \f[B]func.getattr=newest\f[].
  1417. Remember though that this is just \f[B]stat\f[].
  1418. If the file is later \f[B]open\f[]\[aq]ed or \f[B]unlink\f[]\[aq]ed and
  1419. the policy is different for those then a completely different file or
  1420. directory could be acted on.
  1421. .IP \[bu] 2
  1422. Some policies mixed with some functions may result in strange behaviors.
  1423. Not that some of these behaviors and race conditions couldn\[aq]t happen
  1424. outside \f[B]mergerfs\f[] but that they are far more likely to occur on
  1425. account of the attempt to merge together multiple sources of data which
  1426. could be out of sync due to the different policies.
  1427. .IP \[bu] 2
  1428. For consistency its generally best to set \f[B]category\f[] wide
  1429. policies rather than individual \f[B]func\f[]\[aq]s.
  1430. This will help limit the confusion of tools such as
  1431. rsync (http://linux.die.net/man/1/rsync).
  1432. However, the flexibility is there if needed.
  1433. .SH KNOWN ISSUES / BUGS
  1434. .SS directory mtime is not being updated
  1435. .PP
  1436. Remember that the default policy for \f[C]getattr\f[] is \f[C]ff\f[].
  1437. The information for the first directory found will be returned.
  1438. If it wasn\[aq]t the directory which had been updated then it will
  1439. appear outdated.
  1440. .PP
  1441. The reason this is the default is because any other policy would be more
  1442. expensive and for many applications it is unnecessary.
  1443. To always return the directory with the most recent mtime or a faked
  1444. value based on all found would require a scan of all drives.
  1445. .PP
  1446. If you always want the directory information from the one with the most
  1447. recent mtime then use the \f[C]newest\f[] policy for \f[C]getattr\f[].
  1448. .SS \f[C]mv\ /mnt/pool/foo\ /mnt/disk1/foo\f[] removes \f[C]foo\f[]
  1449. .PP
  1450. This is not a bug.
  1451. .PP
  1452. Run in verbose mode to better undertand what\[aq]s happening:
  1453. .IP
  1454. .nf
  1455. \f[C]
  1456. $\ mv\ \-v\ /mnt/pool/foo\ /mnt/disk1/foo
  1457. copied\ \[aq]/mnt/pool/foo\[aq]\ \->\ \[aq]/mnt/disk1/foo\[aq]
  1458. removed\ \[aq]/mnt/pool/foo\[aq]
  1459. $\ ls\ /mnt/pool/foo
  1460. ls:\ cannot\ access\ \[aq]/mnt/pool/foo\[aq]:\ No\ such\ file\ or\ directory
  1461. \f[]
  1462. .fi
  1463. .PP
  1464. \f[C]mv\f[], when working across devices, is copying the source to
  1465. target and then removing the source.
  1466. Since the source \f[B]is\f[] the target in this case, depending on the
  1467. unlink policy, it will remove the just copied file and other files
  1468. across the branches.
  1469. .PP
  1470. If you want to move files to one drive just copy them there and use
  1471. mergerfs.dedup to clean up the old paths or manually remove them from
  1472. the branches directly.
  1473. .SS cached memory appears greater than it should be
  1474. .PP
  1475. Use \f[C]cache.files=off\f[] or \f[C]direct_io=true\f[].
  1476. See the section on page caching.
  1477. .SS NFS clients returning ESTALE / Stale file handle
  1478. .PP
  1479. Be sure to use \f[C]noforget\f[] and \f[C]use_ino\f[] arguments.
  1480. .SS NFS clients don\[aq]t work
  1481. .PP
  1482. Some NFS clients appear to fail when a mergerfs mount is exported.
  1483. Kodi in particular seems to have issues.
  1484. .PP
  1485. Try enabling the \f[C]use_ino\f[] option.
  1486. Some have reported that it fixes the issue.
  1487. .SS rtorrent fails with ENODEV (No such device)
  1488. .PP
  1489. Be sure to set \f[C]cache.files=partial|full|auto\-full\f[] or turn off
  1490. \f[C]direct_io\f[].
  1491. rtorrent and some other applications use
  1492. mmap (http://linux.die.net/man/2/mmap) to read and write to files and
  1493. offer no failback to traditional methods.
  1494. FUSE does not currently support mmap while using \f[C]direct_io\f[].
  1495. There may be a performance penalty on writes with \f[C]direct_io\f[] off
  1496. as well as the problem of double caching but it\[aq]s the only way to
  1497. get such applications to work.
  1498. If the performance loss is too high for other apps you can mount
  1499. mergerfs twice.
  1500. Once with \f[C]direct_io\f[] enabled and one without it.
  1501. Be sure to set \f[C]dropcacheonclose=true\f[] if not using
  1502. \f[C]direct_io\f[].
  1503. .SS rtorrent fails with files >= 4GiB
  1504. .PP
  1505. This is a kernel bug with mmap and FUSE on 32bit platforms.
  1506. A fix should become available for all LTS releases.
  1507. .PP
  1508. https://marc.info/?l=linux\-fsdevel&m=155550785230874&w=2
  1509. .SS Crashing on OpenVZ
  1510. .PP
  1511. There appears to be a bug in the OpenVZ kernel with regard to how it
  1512. handles ioctl calls.
  1513. It is making invalid requests which leads to a crash.
  1514. As of 2019\-12\-10 there is a bug report filed with OpenVZ but it is not
  1515. yet fixed.
  1516. .SS Plex doesn\[aq]t work with mergerfs
  1517. .PP
  1518. It does.
  1519. If you\[aq]re trying to put Plex\[aq]s config / metadata on mergerfs you
  1520. have to leave \f[C]direct_io\f[] off because Plex is using sqlite which
  1521. apparently needs mmap.
  1522. mmap doesn\[aq]t work with \f[C]direct_io\f[].
  1523. To fix this place the data elsewhere or disable \f[C]direct_io\f[] (with
  1524. \f[C]dropcacheonclose=true\f[]).
  1525. .PP
  1526. If the issue is that scanning doesn\[aq]t seem to pick up media then be
  1527. sure to set \f[C]func.getattr=newest\f[] as mentioned above.
  1528. .SS mmap performance is really bad
  1529. .PP
  1530. There is a bug (https://lkml.org/lkml/2016/3/16/260) in caching which
  1531. affects overall performance of mmap through FUSE in Linux 4.x kernels.
  1532. It is fixed in 4.4.10 and 4.5.4 (https://lkml.org/lkml/2016/5/11/59).
  1533. .SS When a program tries to move or rename a file it fails
  1534. .PP
  1535. Please read the section above regarding rename & link (#rename--link).
  1536. .PP
  1537. The problem is that many applications do not properly handle
  1538. \f[C]EXDEV\f[] errors which \f[C]rename\f[] and \f[C]link\f[] may return
  1539. even though they are perfectly valid situations which do not indicate
  1540. actual drive or OS errors.
  1541. The error will only be returned by mergerfs if using a path preserving
  1542. policy as described in the policy section above.
  1543. If you do not care about path preservation simply change the mergerfs
  1544. policy to the non\-path preserving version.
  1545. For example: \f[C]\-o\ category.create=mfs\f[]
  1546. .PP
  1547. Ideally the offending software would be fixed and it is recommended that
  1548. if you run into this problem you contact the software\[aq]s author and
  1549. request proper handling of \f[C]EXDEV\f[] errors.
  1550. .SS Samba: Moving files / directories fails
  1551. .PP
  1552. Workaround: Copy the file/directory and then remove the original rather
  1553. than move.
  1554. .PP
  1555. This isn\[aq]t an issue with Samba but some SMB clients.
  1556. GVFS\-fuse v1.20.3 and prior (found in Ubuntu 14.04 among others) failed
  1557. to handle certain error codes correctly.
  1558. Particularly \f[B]STATUS_NOT_SAME_DEVICE\f[] which comes from the
  1559. \f[B]EXDEV\f[] which is returned by \f[B]rename\f[] when the call is
  1560. crossing mount points.
  1561. When a program gets an \f[B]EXDEV\f[] it needs to explicitly take an
  1562. alternate action to accomplish its goal.
  1563. In the case of \f[B]mv\f[] or similar it tries \f[B]rename\f[] and on
  1564. \f[B]EXDEV\f[] falls back to a manual copying of data between the two
  1565. locations and unlinking the source.
  1566. In these older versions of GVFS\-fuse if it received \f[B]EXDEV\f[] it
  1567. would translate that into \f[B]EIO\f[].
  1568. This would cause \f[B]mv\f[] or most any application attempting to move
  1569. files around on that SMB share to fail with a IO error.
  1570. .PP
  1571. GVFS\-fuse v1.22.0 (https://bugzilla.gnome.org/show_bug.cgi?id=734568)
  1572. and above fixed this issue but a large number of systems use the older
  1573. release.
  1574. On Ubuntu the version can be checked by issuing
  1575. \f[C]apt\-cache\ showpkg\ gvfs\-fuse\f[].
  1576. Most distros released in 2015 seem to have the updated release and will
  1577. work fine but older systems may not.
  1578. Upgrading gvfs\-fuse or the distro in general will address the problem.
  1579. .PP
  1580. In Apple\[aq]s MacOSX 10.9 they replaced Samba (client and server) with
  1581. their own product.
  1582. It appears their new client does not handle \f[B]EXDEV\f[] either and
  1583. responds similar to older release of gvfs on Linux.
  1584. .SS Trashing files occasionally fails
  1585. .PP
  1586. This is the same issue as with Samba.
  1587. \f[C]rename\f[] returns \f[C]EXDEV\f[] (in our case that will really
  1588. only happen with path preserving policies like \f[C]epmfs\f[]) and the
  1589. software doesn\[aq]t handle the situtation well.
  1590. This is unfortunately a common failure of software which moves files
  1591. around.
  1592. The standard indicates that an implementation \f[C]MAY\f[] choose to
  1593. support non\-user home directory trashing of files (which is a
  1594. \f[C]MUST\f[]).
  1595. The implementation \f[C]MAY\f[] also support "top directory trashes"
  1596. which many probably do.
  1597. .PP
  1598. To create a \f[C]$topdir/.Trash\f[] directory as defined in the standard
  1599. use the mergerfs\-tools (https://github.com/trapexit/mergerfs-tools)
  1600. tool \f[C]mergerfs.mktrash\f[].
  1601. .SS tar: Directory renamed before its status could be extracted
  1602. .PP
  1603. Make sure to use the \f[C]use_ino\f[] option.
  1604. .SS Supplemental user groups
  1605. .PP
  1606. Due to the overhead of
  1607. getgroups/setgroups (http://linux.die.net/man/2/setgroups) mergerfs
  1608. utilizes a cache.
  1609. This cache is opportunistic and per thread.
  1610. Each thread will query the supplemental groups for a user when that
  1611. particular thread needs to change credentials and will keep that data
  1612. for the lifetime of the thread.
  1613. This means that if a user is added to a group it may not be picked up
  1614. without the restart of mergerfs.
  1615. However, since the high level FUSE API\[aq]s (at least the standard
  1616. version) thread pool dynamically grows and shrinks it\[aq]s possible
  1617. that over time a thread will be killed and later a new thread with no
  1618. cache will start and query the new data.
  1619. .PP
  1620. The gid cache uses fixed storage to simplify the design and be
  1621. compatible with older systems which may not have C++11 compilers.
  1622. There is enough storage for 256 users\[aq] supplemental groups.
  1623. Each user is allowed upto 32 supplemental groups.
  1624. Linux >= 2.6.3 allows upto 65535 groups per user but most other *nixs
  1625. allow far less.
  1626. NFS allowing only 16.
  1627. The system does handle overflow gracefully.
  1628. If the user has more than 32 supplemental groups only the first 32 will
  1629. be used.
  1630. If more than 256 users are using the system when an uncached user is
  1631. found it will evict an existing user\[aq]s cache at random.
  1632. So long as there aren\[aq]t more than 256 active users this should be
  1633. fine.
  1634. If either value is too low for your needs you will have to modify
  1635. \f[C]gidcache.hpp\f[] to increase the values.
  1636. Note that doing so will increase the memory needed by each thread.
  1637. .SS mergerfs or libfuse crashing
  1638. .PP
  1639. \f[B]NOTE:\f[] as of mergerfs 2.22.0 it includes the most recent version
  1640. of libfuse (or requires libfuse\-2.9.7) so any crash should be reported.
  1641. For older releases continue reading...
  1642. .PP
  1643. If suddenly the mergerfs mount point disappears and
  1644. \f[C]Transport\ endpoint\ is\ not\ connected\f[] is returned when
  1645. attempting to perform actions within the mount directory \f[B]and\f[]
  1646. the version of libfuse (use \f[C]mergerfs\ \-v\f[] to find the version)
  1647. is older than \f[C]2.9.4\f[] its likely due to a bug in libfuse.
  1648. Affected versions of libfuse can be found in Debian Wheezy, Ubuntu
  1649. Precise and others.
  1650. .PP
  1651. In order to fix this please install newer versions of libfuse.
  1652. If using a Debian based distro (Debian,Ubuntu,Mint) you can likely just
  1653. install newer versions of
  1654. libfuse (https://packages.debian.org/unstable/libfuse2) and
  1655. fuse (https://packages.debian.org/unstable/fuse) from the repo of a
  1656. newer release.
  1657. .SS mergerfs appears to be crashing or exiting
  1658. .PP
  1659. There seems to be an issue with Linux version \f[C]4.9.0\f[] and above
  1660. in which an invalid message appears to be transmitted to libfuse (used
  1661. by mergerfs) causing it to exit.
  1662. No messages will be printed in any logs as its not a proper crash.
  1663. Debugging of the issue is still ongoing and can be followed via the
  1664. fuse\-devel
  1665. thread (https://sourceforge.net/p/fuse/mailman/message/35662577).
  1666. .SS mergerfs under heavy load and memory preasure leads to kernel panic
  1667. .PP
  1668. https://lkml.org/lkml/2016/9/14/527
  1669. .IP
  1670. .nf
  1671. \f[C]
  1672. [25192.515454]\ kernel\ BUG\ at\ /build/linux\-a2WvEb/linux\-4.4.0/mm/workingset.c:346!
  1673. [25192.517521]\ invalid\ opcode:\ 0000\ [#1]\ SMP
  1674. [25192.519602]\ Modules\ linked\ in:\ netconsole\ ip6t_REJECT\ nf_reject_ipv6\ ipt_REJECT\ nf_reject_ipv4\ configfs\ binfmt_misc\ veth\ bridge\ stp\ llc\ nf_conntrack_ipv6\ nf_defrag_ipv6\ xt_conntrack\ ip6table_filter\ ip6_tables\ xt_multiport\ iptable_filter\ ipt_MASQUERADE\ nf_nat_masquerade_ipv4\ xt_comment\ xt_nat\ iptable_nat\ nf_conntrack_ipv4\ nf_defrag_ipv4\ nf_nat_ipv4\ nf_nat\ nf_conntrack\ xt_CHECKSUM\ xt_tcpudp\ iptable_mangle\ ip_tables\ x_tables\ intel_rapl\ x86_pkg_temp_thermal\ intel_powerclamp\ eeepc_wmi\ asus_wmi\ coretemp\ sparse_keymap\ kvm_intel\ ppdev\ kvm\ irqbypass\ mei_me\ 8250_fintek\ input_leds\ serio_raw\ parport_pc\ tpm_infineon\ mei\ shpchp\ mac_hid\ parport\ lpc_ich\ autofs4\ drbg\ ansi_cprng\ dm_crypt\ algif_skcipher\ af_alg\ btrfs\ raid456\ async_raid6_recov\ async_memcpy\ async_pq\ async_xor\ async_tx\ xor\ raid6_pq\ libcrc32c\ raid0\ multipath\ linear\ raid10\ raid1\ i915\ crct10dif_pclmul\ crc32_pclmul\ aesni_intel\ i2c_algo_bit\ aes_x86_64\ drm_kms_helper\ lrw\ gf128mul\ glue_helper\ ablk_helper\ syscopyarea\ cryptd\ sysfillrect\ sysimgblt\ fb_sys_fops\ drm\ ahci\ r8169\ libahci\ mii\ wmi\ fjes\ video\ [last\ unloaded:\ netconsole]
  1675. [25192.540910]\ CPU:\ 2\ PID:\ 63\ Comm:\ kswapd0\ Not\ tainted\ 4.4.0\-36\-generic\ #55\-Ubuntu
  1676. [25192.543411]\ Hardware\ name:\ System\ manufacturer\ System\ Product\ Name/P8H67\-M\ PRO,\ BIOS\ 3904\ 04/27/2013
  1677. [25192.545840]\ task:\ ffff88040cae6040\ ti:\ ffff880407488000\ task.ti:\ ffff880407488000
  1678. [25192.548277]\ RIP:\ 0010:[<ffffffff811ba501>]\ \ [<ffffffff811ba501>]\ shadow_lru_isolate+0x181/0x190
  1679. [25192.550706]\ RSP:\ 0018:ffff88040748bbe0\ \ EFLAGS:\ 00010002
  1680. [25192.553127]\ RAX:\ 0000000000001c81\ RBX:\ ffff8802f91ee928\ RCX:\ ffff8802f91eeb38
  1681. [25192.555544]\ RDX:\ ffff8802f91ee938\ RSI:\ ffff8802f91ee928\ RDI:\ ffff8804099ba2c0
  1682. [25192.557914]\ RBP:\ ffff88040748bc08\ R08:\ 000000000001a7b6\ R09:\ 000000000000003f
  1683. [25192.560237]\ R10:\ 000000000001a750\ R11:\ 0000000000000000\ R12:\ ffff8804099ba2c0
  1684. [25192.562512]\ R13:\ ffff8803157e9680\ R14:\ ffff8803157e9668\ R15:\ ffff8804099ba2c8
  1685. [25192.564724]\ FS:\ \ 0000000000000000(0000)\ GS:ffff88041f280000(0000)\ knlGS:0000000000000000
  1686. [25192.566990]\ CS:\ \ 0010\ DS:\ 0000\ ES:\ 0000\ CR0:\ 0000000080050033
  1687. [25192.569201]\ CR2:\ 00007ffabb690000\ CR3:\ 0000000001e0a000\ CR4:\ 00000000000406e0
  1688. [25192.571419]\ Stack:
  1689. [25192.573550]\ \ ffff8804099ba2c0\ ffff88039e4f86f0\ ffff8802f91ee928\ ffff8804099ba2c8
  1690. [25192.575695]\ \ ffff88040748bd08\ ffff88040748bc58\ ffffffff811b99bf\ 0000000000000052
  1691. [25192.577814]\ \ 0000000000000000\ ffffffff811ba380\ 000000000000008a\ 0000000000000080
  1692. [25192.579947]\ Call\ Trace:
  1693. [25192.582022]\ \ [<ffffffff811b99bf>]\ __list_lru_walk_one.isra.3+0x8f/0x130
  1694. [25192.584137]\ \ [<ffffffff811ba380>]\ ?\ memcg_drain_all_list_lrus+0x190/0x190
  1695. [25192.586165]\ \ [<ffffffff811b9a83>]\ list_lru_walk_one+0x23/0x30
  1696. [25192.588145]\ \ [<ffffffff811ba544>]\ scan_shadow_nodes+0x34/0x50
  1697. [25192.590074]\ \ [<ffffffff811a0e9d>]\ shrink_slab.part.40+0x1ed/0x3d0
  1698. [25192.591985]\ \ [<ffffffff811a53da>]\ shrink_zone+0x2ca/0x2e0
  1699. [25192.593863]\ \ [<ffffffff811a64ce>]\ kswapd+0x51e/0x990
  1700. [25192.595737]\ \ [<ffffffff811a5fb0>]\ ?\ mem_cgroup_shrink_node_zone+0x1c0/0x1c0
  1701. [25192.597613]\ \ [<ffffffff810a0808>]\ kthread+0xd8/0xf0
  1702. [25192.599495]\ \ [<ffffffff810a0730>]\ ?\ kthread_create_on_node+0x1e0/0x1e0
  1703. [25192.601335]\ \ [<ffffffff8182e34f>]\ ret_from_fork+0x3f/0x70
  1704. [25192.603193]\ \ [<ffffffff810a0730>]\ ?\ kthread_create_on_node+0x1e0/0x1e0
  1705. \f[]
  1706. .fi
  1707. .PP
  1708. There is a bug in the kernel.
  1709. A work around appears to be turning off \f[C]splice\f[].
  1710. Don\[aq]t add the \f[C]splice_*\f[] arguments or add
  1711. \f[C]no_splice_write,no_splice_move,no_splice_read\f[].
  1712. This, however, is not guaranteed to work.
  1713. .SS rm: fts_read failed: No such file or directory
  1714. .PP
  1715. NOTE: This is only relevant to mergerfs versions at or below v2.25.x and
  1716. should not occur in more recent versions.
  1717. See the notes on \f[C]unlink\f[].
  1718. .PP
  1719. Not \f[I]really\f[] a bug.
  1720. The FUSE library will move files when asked to delete them as a way to
  1721. deal with certain edge cases and then later delete that file when its
  1722. clear the file is no longer needed.
  1723. This however can lead to two issues.
  1724. One is that these hidden files are noticed by \f[C]rm\ \-rf\f[] or
  1725. \f[C]find\f[] when scanning directories and they may try to remove them
  1726. and they might have disappeared already.
  1727. There is nothing \f[I]wrong\f[] about this happening but it can be
  1728. annoying.
  1729. The second issue is that a directory might not be able to removed on
  1730. account of the hidden file being still there.
  1731. .PP
  1732. Using the \f[B]hard_remove\f[] option will make it so these temporary
  1733. files are not used and files are deleted immedately.
  1734. That has a side effect however.
  1735. Files which are unlinked and then they are still used (in certain forms)
  1736. will result in an error (ENOENT).
  1737. .SH FAQ
  1738. .SS How well does mergerfs scale? Is it "production ready?"
  1739. .PP
  1740. Users have reported running mergerfs on everything from a Raspberry Pi
  1741. to dual socket Xeon systems with >20 cores.
  1742. I\[aq]m aware of at least a few companies which use mergerfs in
  1743. production.
  1744. Open Media Vault (https://www.openmediavault.org) includes mergerfs as
  1745. its sole solution for pooling drives.
  1746. .SS Can mergerfs be used with drives which already have data / are in
  1747. use?
  1748. .PP
  1749. Yes.
  1750. MergerFS is a proxy and does \f[B]NOT\f[] interfere with the normal form
  1751. or function of the drives / mounts / paths it manages.
  1752. .PP
  1753. MergerFS is \f[B]not\f[] a traditional filesystem.
  1754. MergerFS is \f[B]not\f[] RAID.
  1755. It does \f[B]not\f[] manipulate the data that passes through it.
  1756. It does \f[B]not\f[] shard data across drives.
  1757. It merely shards some \f[B]behavior\f[] and aggregates others.
  1758. .SS Can mergerfs be removed without affecting the data?
  1759. .PP
  1760. See the previous question\[aq]s answer.
  1761. .SS Do hard links work?
  1762. .PP
  1763. Yes.
  1764. You need to use \f[C]use_ino\f[] to support proper reporting of inodes.
  1765. .PP
  1766. What mergerfs does not do is fake hard links across branches.
  1767. Read the section "rename & link" for how it works.
  1768. .SS Does mergerfs support CoW / copy\-on\-write?
  1769. .PP
  1770. Not in the sense of a filesystem like BTRFS or ZFS nor in the overlayfs
  1771. or aufs sense.
  1772. It does offer a
  1773. cow\-shell (http://manpages.ubuntu.com/manpages/bionic/man1/cow-shell.1.html)
  1774. like hard link breaking (copy to temp file then rename over original)
  1775. which can be useful when wanting to save space by hardlinking duplicate
  1776. files but wish to treat each name as if it were a unique and separate
  1777. file.
  1778. .SS Why can\[aq]t I see my files / directories?
  1779. .PP
  1780. It\[aq]s almost always a permissions issue.
  1781. Unlike mhddfs, which runs as root and attempts to access content as
  1782. such, mergerfs always changes its credentials to that of the caller.
  1783. This means that if the user does not have access to a file or directory
  1784. than neither will mergerfs.
  1785. However, because mergerfs is creating a union of paths it may be able to
  1786. read some files and directories on one drive but not another resulting
  1787. in an incomplete set.
  1788. .PP
  1789. Whenever you run into a split permission issue (seeing some but not all
  1790. files) try using
  1791. mergerfs.fsck (https://github.com/trapexit/mergerfs-tools) tool to check
  1792. for and fix the mismatch.
  1793. If you aren\[aq]t seeing anything at all be sure that the basic
  1794. permissions are correct.
  1795. The user and group values are correct and that directories have their
  1796. executable bit set.
  1797. A common mistake by users new to Linux is to \f[C]chmod\ \-R\ 644\f[]
  1798. when they should have \f[C]chmod\ \-R\ u=rwX,go=rX\f[].
  1799. .PP
  1800. If using a network filesystem such as NFS, SMB, CIFS (Samba) be sure to
  1801. pay close attention to anything regarding permissioning and users.
  1802. Root squashing and user translation for instance has bitten a few
  1803. mergerfs users.
  1804. Some of these also affect the use of mergerfs from container platforms
  1805. such as Docker.
  1806. .SS Why is only one drive being used?
  1807. .PP
  1808. Are you using a path preserving policy?
  1809. The default policy for file creation is \f[C]epmfs\f[].
  1810. That means only the drives with the path preexisting will be considered
  1811. when creating a file.
  1812. If you don\[aq]t care about where files and directories are created you
  1813. likely shouldn\[aq]t be using a path preserving policy and instead
  1814. something like \f[C]mfs\f[].
  1815. .PP
  1816. This can be especially apparent when filling an empty pool from an
  1817. external source.
  1818. If you do want path preservation you\[aq]ll need to perform the manual
  1819. act of creating paths on the drives you want the data to land on before
  1820. transfering your data.
  1821. Setting \f[C]func.mkdir=epall\f[] can simplify managing path
  1822. perservation for \f[C]create\f[].
  1823. .SS Why was libfuse embedded into mergerfs?
  1824. .IP "1." 3
  1825. A significant number of users use mergerfs on distros with old versions
  1826. of libfuse which have serious bugs.
  1827. Requiring updated versions of libfuse on those distros isn\[aq]t
  1828. pratical (no package offered, user inexperience, etc.).
  1829. The only practical way to provide a stable runtime on those systems was
  1830. to "vendor" / embed the library into the project.
  1831. .IP "2." 3
  1832. mergerfs was written to use the high level API.
  1833. There are a number of limitations in the HLAPI that make certain
  1834. features difficult or impossible to implement.
  1835. While some of these features could be patched into newer versions of
  1836. libfuse without breaking the public API some of them would require hacky
  1837. code to provide backwards compatibility.
  1838. While it may still be worth working with upstream to address these
  1839. issues in future versions, since the library needs to be vendored for
  1840. stability and compatibility reasons it is preferable / easier to modify
  1841. the API.
  1842. Longer term the plan is to rewrite mergerfs to use the low level API.
  1843. .SS Why did support for system libfuse get removed?
  1844. .PP
  1845. See above first.
  1846. .PP
  1847. If/when mergerfs is rewritten to use the low\-level API then it\[aq]ll
  1848. be plausible to support system libfuse but till then its simply too much
  1849. work to manage the differences across the versions.
  1850. .SS Why use mergerfs over mhddfs?
  1851. .PP
  1852. mhddfs is no longer maintained and has some known stability and security
  1853. issues (see below).
  1854. MergerFS provides a superset of mhddfs\[aq] features and should offer
  1855. the same or maybe better performance.
  1856. .PP
  1857. Below is an example of mhddfs and mergerfs setup to work similarly.
  1858. .PP
  1859. \f[C]mhddfs\ \-o\ mlimit=4G,allow_other\ /mnt/drive1,/mnt/drive2\ /mnt/pool\f[]
  1860. .PP
  1861. \f[C]mergerfs\ \-o\ minfreespace=4G,allow_other,category.create=ff\ /mnt/drive1:/mnt/drive2\ /mnt/pool\f[]
  1862. .SS Why use mergerfs over aufs?
  1863. .PP
  1864. aufs is mostly abandoned and no longer available in many distros.
  1865. .PP
  1866. While aufs can offer better peak performance mergerfs provides more
  1867. configurability and is generally easier to use.
  1868. mergerfs however does not offer the overlay / copy\-on\-write (CoW)
  1869. features which aufs and overlayfs have.
  1870. .SS Why use mergerfs over unionfs?
  1871. .PP
  1872. UnionFS is more like aufs than mergerfs in that it offers overlay / CoW
  1873. features.
  1874. If you\[aq]re just looking to create a union of drives and want
  1875. flexibility in file/directory placement then mergerfs offers that
  1876. whereas unionfs is more for overlaying RW filesystems over RO ones.
  1877. .SS Why use mergerfs over LVM/ZFS/BTRFS/RAID0 drive concatenation /
  1878. striping?
  1879. .PP
  1880. With simple JBOD / drive concatenation / stripping / RAID0 a single
  1881. drive failure will result in full pool failure.
  1882. mergerfs performs a similar behavior without the possibility of
  1883. catastrophic failure and the difficulties in recovery.
  1884. Drives may fail however all other data will continue to be accessable.
  1885. .PP
  1886. When combined with something like SnapRaid (http://www.snapraid.it)
  1887. and/or an offsite backup solution you can have the flexibilty of JBOD
  1888. without the single point of failure.
  1889. .SS Why use mergerfs over ZFS?
  1890. .PP
  1891. MergerFS is not intended to be a replacement for ZFS.
  1892. MergerFS is intended to provide flexible pooling of arbitrary drives
  1893. (local or remote), of arbitrary sizes, and arbitrary filesystems.
  1894. For \f[C]write\ once,\ read\ many\f[] usecases such as bulk media
  1895. storage.
  1896. Where data integrity and backup is managed in other ways.
  1897. In that situation ZFS can introduce major maintance and cost burdens as
  1898. described
  1899. here (http://louwrentius.com/the-hidden-cost-of-using-zfs-for-your-home-nas.html).
  1900. .SS Can drives be written to directly? Outside of mergerfs while pooled?
  1901. .PP
  1902. Yes, however its not recommended to use the same file from within the
  1903. pool and from without at the same time.
  1904. Especially if using caching of any kind (cache.files, cache.entry,
  1905. cache.attr, cache.negative_entry, cache.symlinks, cache.readdir, etc.).
  1906. .SS Why do I get an "out of space" / "no space left on device" / ENOSPC
  1907. error even though there appears to be lots of space available?
  1908. .PP
  1909. First make sure you\[aq]ve read the sections above about policies, path
  1910. preservation, branch filtering, and the options \f[B]minfreespace\f[],
  1911. \f[B]moveonenospc\f[], \f[B]statfs\f[], and \f[B]statfs_ignore\f[].
  1912. .PP
  1913. mergerfs is simply presenting a union of the content within multiple
  1914. branches.
  1915. The reported free space is an aggregate of space available within the
  1916. pool (behavior modified by \f[B]statfs\f[] and \f[B]statfs_ignore\f[]).
  1917. It does not represent a contiguous space.
  1918. In the same way that read\-only filesystems, those with quotas, or
  1919. reserved space report the full theoretical space available.
  1920. .PP
  1921. Due to path preservation, branch tagging, read\-only status, and
  1922. \f[B]minfreespace\f[] settings it is perfectly valid that
  1923. \f[C]ENOSPC\f[] / "out of space" / "no space left on device" be
  1924. returned.
  1925. It is doing what was asked of it: filtering possible branches due to
  1926. those settings.
  1927. Only one error can be returned and if one of the reasons for filtering a
  1928. branch was \f[B]minfreespace\f[] then it will be returned as such.
  1929. \f[B]moveonenospc\f[] is only relevant to writing a file which is too
  1930. large for the drive its currently on.
  1931. .PP
  1932. It is also possible that the filesystem selected has run out of inodes.
  1933. Use \f[C]df\ \-i\f[] to list the total and available inodes per
  1934. filesystem.
  1935. .PP
  1936. If you don\[aq]t care about path preservation then simply change the
  1937. \f[C]create\f[] policy to one which isn\[aq]t.
  1938. \f[C]mfs\f[] is probably what most are looking for.
  1939. The reason its not default is because it was originally set to
  1940. \f[C]epmfs\f[] and changing it now would change people\[aq]s setup.
  1941. Such a setting change will likely occur in mergerfs 3.
  1942. .SS Can mergerfs mounts be exported over NFS?
  1943. .PP
  1944. Yes.
  1945. Due to current usage of libfuse by mergerfs and how NFS interacts with
  1946. it it is necessary to add \f[C]noforget\f[] to mergerfs options to keep
  1947. from getting "stale file handle" errors.
  1948. .PP
  1949. Some clients (Kodi) have issues in which the contents of the NFS mount
  1950. will not be presented but users have found that enabling the
  1951. \f[C]use_ino\f[] option often fixes that problem.
  1952. .SS Can mergerfs mounts be exported over Samba / SMB?
  1953. .PP
  1954. Yes.
  1955. While some users have reported problems it appears to always be related
  1956. to how Samba is setup in relation to permissions.
  1957. .SS How are inodes calculated?
  1958. .PP
  1959. mergerfs\-inode = (original\-inode | (device\-id << 32))
  1960. .PP
  1961. While \f[C]ino_t\f[] is 64 bits only a few filesystems use more than 32.
  1962. Similarly, while \f[C]dev_t\f[] is also 64 bits it was traditionally 16
  1963. bits.
  1964. Bitwise or\[aq]ing them together should work most of the time.
  1965. While totally unique inodes are preferred the overhead which would be
  1966. needed does not seem to outweighted by the benefits.
  1967. .PP
  1968. While atypical, yes, inodes can be reused and not refer to the same
  1969. file.
  1970. The internal id used to reference a file in FUSE is different from the
  1971. inode value presented.
  1972. The former is the \f[C]nodeid\f[] and is actually a tuple of
  1973. (nodeid,generation).
  1974. That tuple is not user facing.
  1975. The inode is merely metadata passed through the kernel and found using
  1976. the \f[C]stat\f[] family of calls or \f[C]readdir\f[].
  1977. .PP
  1978. From FUSE docs regarding \f[C]use_ino\f[]:
  1979. .IP
  1980. .nf
  1981. \f[C]
  1982. Honor\ the\ st_ino\ field\ in\ the\ functions\ getattr()\ and
  1983. fill_dir().\ This\ value\ is\ used\ to\ fill\ in\ the\ st_ino\ field
  1984. in\ the\ stat(2),\ lstat(2),\ fstat(2)\ functions\ and\ the\ d_ino
  1985. field\ in\ the\ readdir(2)\ function.\ The\ filesystem\ does\ not
  1986. have\ to\ guarantee\ uniqueness,\ however\ some\ applications
  1987. rely\ on\ this\ value\ being\ unique\ for\ the\ whole\ filesystem.
  1988. Note\ that\ this\ does\ *not*\ affect\ the\ inode\ that\ libfuse
  1989. and\ the\ kernel\ use\ internally\ (also\ called\ the\ "nodeid").
  1990. \f[]
  1991. .fi
  1992. .SS I notice massive slowdowns of writes over NFS
  1993. .PP
  1994. Due to how NFS works and interacts with FUSE when not using
  1995. \f[C]cache.files=off\f[] or \f[C]direct_io\f[] its possible that a
  1996. getxattr for \f[C]security.capability\f[] will be issued prior to any
  1997. write.
  1998. This will usually result in a massive slowdown for writes.
  1999. Using \f[C]cache.files=off\f[] or \f[C]direct_io\f[] will keep this from
  2000. happening (and generally good to enable unless you need the features it
  2001. disables) but the \f[C]security_capability\f[] option can also help by
  2002. short circuiting the call and returning \f[C]ENOATTR\f[].
  2003. .PP
  2004. You could also set \f[C]xattr\f[] to \f[C]noattr\f[] or \f[C]nosys\f[]
  2005. to short circuit or stop all xattr requests.
  2006. .SS What are these .fuse_hidden files?
  2007. .PP
  2008. NOTE: mergerfs >= 2.26.0 will not have these temporary files.
  2009. See the notes on \f[C]unlink\f[].
  2010. .PP
  2011. When not using \f[B]hard_remove\f[] libfuse will create
  2012. \&.fuse_hiddenXXXXXXXX files when an opened file is unlinked.
  2013. This is to simplify "use after unlink" usecases.
  2014. There is a possibility these files end up being picked up by software
  2015. scanning directories and not ignoring hidden files.
  2016. This is rarely a problem but a solution is in the works.
  2017. .PP
  2018. The files are cleaned up once the file is finally closed.
  2019. Only if mergerfs crashes or is killed would they be left around.
  2020. They are safe to remove as they are already unlinked files.
  2021. .SS It\[aq]s mentioned that there are some security issues with mhddfs.
  2022. What are they? How does mergerfs address them?
  2023. .PP
  2024. mhddfs (https://github.com/trapexit/mhddfs) manages running as
  2025. \f[B]root\f[] by calling
  2026. getuid() (https://github.com/trapexit/mhddfs/blob/cae96e6251dd91e2bdc24800b4a18a74044f6672/src/main.c#L319)
  2027. and if it returns \f[B]0\f[] then it will
  2028. chown (http://linux.die.net/man/1/chown) the file.
  2029. Not only is that a race condition but it doesn\[aq]t handle other
  2030. situations.
  2031. Rather than attempting to simulate POSIX ACL behavior the proper way to
  2032. manage this is to use seteuid (http://linux.die.net/man/2/seteuid) and
  2033. setegid (http://linux.die.net/man/2/setegid), in effect becoming the
  2034. user making the original call, and perform the action as them.
  2035. This is what mergerfs does and why mergerfs should always run as root.
  2036. .PP
  2037. In Linux setreuid syscalls apply only to the thread.
  2038. GLIBC hides this away by using realtime signals to inform all threads to
  2039. change credentials.
  2040. Taking after \f[B]Samba\f[], mergerfs uses
  2041. \f[B]syscall(SYS_setreuid,...)\f[] to set the callers credentials for
  2042. that thread only.
  2043. Jumping back to \f[B]root\f[] as necessary should escalated privileges
  2044. be needed (for instance: to clone paths between drives).
  2045. .PP
  2046. For non\-Linux systems mergerfs uses a read\-write lock and changes
  2047. credentials only when necessary.
  2048. If multiple threads are to be user X then only the first one will need
  2049. to change the processes credentials.
  2050. So long as the other threads need to be user X they will take a readlock
  2051. allowing multiple threads to share the credentials.
  2052. Once a request comes in to run as user Y that thread will attempt a
  2053. write lock and change to Y\[aq]s credentials when it can.
  2054. If the ability to give writers priority is supported then that flag will
  2055. be used so threads trying to change credentials don\[aq]t starve.
  2056. This isn\[aq]t the best solution but should work reasonably well
  2057. assuming there are few users.
  2058. .SH PERFORMANCE TWEAKING
  2059. .PP
  2060. NOTE: be sure to read about these features before changing them
  2061. .IP \[bu] 2
  2062. enable (or disable) \f[C]splice_move\f[], \f[C]splice_read\f[], and
  2063. \f[C]splice_write\f[]
  2064. .IP \[bu] 2
  2065. increase cache timeouts \f[C]cache.attr\f[], \f[C]cache.entry\f[],
  2066. \f[C]cache.negative_entry\f[]
  2067. .IP \[bu] 2
  2068. enable (or disable) page caching (\f[C]cache.files\f[])
  2069. .IP \[bu] 2
  2070. enable \f[C]cache.open\f[]
  2071. .IP \[bu] 2
  2072. enable \f[C]cache.statfs\f[]
  2073. .IP \[bu] 2
  2074. enable \f[C]cache.symlinks\f[]
  2075. .IP \[bu] 2
  2076. enable \f[C]cache.readdir\f[]
  2077. .IP \[bu] 2
  2078. change the number of worker threads
  2079. .IP \[bu] 2
  2080. disable \f[C]security_capability\f[] and/or \f[C]xattr\f[]
  2081. .IP \[bu] 2
  2082. disable \f[C]posix_acl\f[]
  2083. .IP \[bu] 2
  2084. disable \f[C]async_read\f[]
  2085. .IP \[bu] 2
  2086. test theoretical performance using \f[C]nullrw\f[] or mounting a ram
  2087. disk
  2088. .IP \[bu] 2
  2089. use \f[C]symlinkify\f[] if your data is largely static
  2090. .IP \[bu] 2
  2091. use tiered cache drives
  2092. .IP \[bu] 2
  2093. use lvm and lvm cache to place a SSD in front of your HDDs (howto
  2094. coming)
  2095. .SH SUPPORT
  2096. .PP
  2097. Filesystems are very complex and difficult to debug.
  2098. mergerfs, while being just a proxy of sorts, is also very difficult to
  2099. debug given the large number of possible settings it can have itself and
  2100. the massive number of environments it can run in.
  2101. When reporting on a suspected issue \f[B]please, please\f[] include as
  2102. much of the below information as possible otherwise it will be difficult
  2103. or impossible to diagnose.
  2104. Also please make sure to read all of the above documentation as it
  2105. includes nearly every known system or user issue previously encountered.
  2106. .SS Information to include in bug reports
  2107. .IP \[bu] 2
  2108. Version of mergerfs: \f[C]mergerfs\ \-V\f[]
  2109. .IP \[bu] 2
  2110. mergerfs settings: from \f[C]/etc/fstab\f[] or command line execution
  2111. .IP \[bu] 2
  2112. Version of Linux: \f[C]uname\ \-a\f[]
  2113. .IP \[bu] 2
  2114. Versions of any additional software being used
  2115. .IP \[bu] 2
  2116. List of drives, their filesystems, and sizes (before and after issue):
  2117. \f[C]df\ \-h\f[]
  2118. .IP \[bu] 2
  2119. A \f[C]strace\f[] of the app having problems:
  2120. .IP \[bu] 2
  2121. \f[C]strace\ \-f\ \-o\ /tmp/app.strace.txt\ <cmd>\f[]
  2122. .IP \[bu] 2
  2123. A \f[C]strace\f[] of mergerfs while the program is trying to do whatever
  2124. it\[aq]s failing to do:
  2125. .IP \[bu] 2
  2126. \f[C]strace\ \-f\ \-p\ <mergerfsPID>\ \-o\ /tmp/mergerfs.strace.txt\f[]
  2127. .IP \[bu] 2
  2128. \f[B]Precise\f[] directions on replicating the issue.
  2129. Do not leave \f[B]anything\f[] out.
  2130. .IP \[bu] 2
  2131. Try to recreate the problem in the simplist way using standard programs.
  2132. .SS Contact / Issue submission
  2133. .IP \[bu] 2
  2134. github.com: https://github.com/trapexit/mergerfs/issues
  2135. .IP \[bu] 2
  2136. email: trapexit\@spawn.link
  2137. .IP \[bu] 2
  2138. twitter: https://twitter.com/_trapexit
  2139. .IP \[bu] 2
  2140. reddit: https://www.reddit.com/user/trapexit
  2141. .IP \[bu] 2
  2142. discord: https://discord.gg/MpAr69V
  2143. .SS Support development
  2144. .PP
  2145. This software is free to use and released under a very liberal license.
  2146. That said if you like this software and would like to support its
  2147. development donations are welcome.
  2148. .IP \[bu] 2
  2149. PayPal: https://paypal.me/trapexit
  2150. .IP \[bu] 2
  2151. GitHub Sponsors: https://github.com/sponsors/trapexit
  2152. .IP \[bu] 2
  2153. Patreon: https://www.patreon.com/trapexit
  2154. .IP \[bu] 2
  2155. SubscribeStar: https://www.subscribestar.com/trapexit
  2156. .IP \[bu] 2
  2157. Ko\-Fi: https://ko\-fi.com/trapexit
  2158. .IP \[bu] 2
  2159. Bitcoin (BTC): 12CdMhEPQVmjz3SSynkAEuD5q9JmhTDCZA
  2160. .IP \[bu] 2
  2161. Bitcoin Cash (BCH): 1AjPqZZhu7GVEs6JFPjHmtsvmDL4euzMzp
  2162. .IP \[bu] 2
  2163. Ethereum (ETH): 0x09A166B11fCC127324C7fc5f1B572255b3046E94
  2164. .IP \[bu] 2
  2165. Litecoin (LTC): LXAsq6yc6zYU3EbcqyWtHBrH1Ypx4GjUjm
  2166. .IP \[bu] 2
  2167. ZCoin (XZC): a8L5Vz35KdCQe7Y7urK2pcCGau7JsqZ5Gw
  2168. .SH LINKS
  2169. .IP \[bu] 2
  2170. https://spawn.link
  2171. .IP \[bu] 2
  2172. https://github.com/trapexit/mergerfs
  2173. .IP \[bu] 2
  2174. https://github.com/trapexit/mergerfs\-tools
  2175. .IP \[bu] 2
  2176. https://github.com/trapexit/scorch
  2177. .IP \[bu] 2
  2178. https://github.com/trapexit/bbf
  2179. .IP \[bu] 2
  2180. https://github.com/trapexit/backup\-and\-recovery\-howtos
  2181. .SH AUTHORS
  2182. Antonio SJ Musumeci <trapexit@spawn.link>.