1 .\" Copyright (c) 2007 The DragonFly Project. All rights reserved.
3 .\" This code is derived from software contributed to The DragonFly Project
4 .\" by Matthew Dillon <dillon@backplane.com>
6 .\" Redistribution and use in source and binary forms, with or without
7 .\" modification, are permitted provided that the following conditions
10 .\" 1. Redistributions of source code must retain the above copyright
11 .\" notice, this list of conditions and the following disclaimer.
12 .\" 2. Redistributions in binary form must reproduce the above copyright
13 .\" notice, this list of conditions and the following disclaimer in
14 .\" the documentation and/or other materials provided with the
16 .\" 3. Neither the name of The DragonFly Project nor the names of its
17 .\" contributors may be used to endorse or promote products derived
18 .\" from this software without specific, prior written permission.
20 .\" THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
21 .\" ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
22 .\" LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
23 .\" FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
24 .\" COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
25 .\" INCIDENTAL, SPECIAL, EXEMPLARY OR CONSEQUENTIAL DAMAGES (INCLUDING,
26 .\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
27 .\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
28 .\" AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
29 .\" OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
30 .\" OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33 .\" $DragonFly: src/sbin/hammer/hammer.8,v 1.38.2.6 2008/09/25 01:39:33 dillon Exp $
39 .Nd HAMMER file system utility
45 .Op Fl f Ar blkdev[:blkdev]*
46 .\" .Op Fl s Ar linkpath
52 This manual page documents the
54 utility which provides miscellaneous functions related to managing a
57 For a general introduction to the
59 file system, its features, and
60 examples on how to set up and maintain one, see
63 The options are as follows:
64 .Bl -tag -width indent
68 Tell the mirror commands to use a 2-way protocol, which allows
69 automatic negotiation of transaction id ranges. This option is
70 automatically enabled by the
74 Specify recursion for those commands which support it.
76 Specify a bandwidth limit in bytes per second for mirroring streams.
77 This option is typically used to prevent batch mirroring operations from
78 loading down the machine.
79 The bandwidth may be suffixed with
85 values in kilobytes, megabytes, and gigabytes per second.
87 When pruning and reblocking you can instruction
90 object id stored in the specified file.
91 If the file does not exist
93 will start at the beginning.
97 specific period of time and is unable to complete the operation it will
98 write out the current object id so the next run can pick up where it left
102 runs to completion it will delete the cyclefile.
103 .It Fl f Ar blkdev[:blkdev]*
104 Specify the volumes making up a
107 .\" .It Fl s Ar linkpath
108 .\" When pruning a filesystem you can instruct
109 .\" .Nm to create softlinks
110 .\" to available snapshots.
112 When maintaining a streaming mirroring this option specifies the
113 minimum delay after a batch ends before the next batch is allowed
115 The default is five seconds.
117 When pruning and reblocking you can tell the utility to stop after a
118 certain period of time. This option is used along with the cycle file
119 option to prune or reblock a portion of the file system incrementally.
121 Increase verboseness. May be specified multiple times.
124 The commands are as follows:
125 .Bl -tag -width indent
126 .\" ==== synctid ====
127 .It Ar synctid Ar filesystem Op quick
128 Generates a guaranteed, formal 64 bit transaction id representing the
129 current state of the specified
131 file system. The file system will
132 be synced to the media.
136 keyword is specified the file system will be soft-synced, meaning that a
137 crash might still undo the state of the file system as of the transaction
138 id returned but any new modifications will occur after the returned
139 transaction id as expected.
141 .It Ar bstats Op interval
144 B-tree statistics until interrupted.
147 seconds between each display.
148 The default interval is one second.
149 .\" ==== iostats ====
150 .It Ar iostats Op interval
153 I/O statistics until interrupted.
156 seconds between each display.
157 The default interval is one second.
158 .\" ==== history ====
159 .It Ar history Ar path ...
160 Show the modification history for
162 file's inode and data.
165 Dump the B-tree. This command needs the
169 .\" Dump the B-tree, record, large-data, and small-data blockmaps, showing
170 .\" physical block assignments and free space percentages.
171 .\" ==== namekey ====
172 .It Ar namekey Ar filename
175 64 bit directory hash for the specified file name.
176 The low 32 bits are used as an iterator for hash collisions and will be
178 .\" ==== namekey32 ====
179 .It Ar namekey32 Ar filename
180 Generate the top 32 bits of a
182 64 bit directory hash for the specified
184 .\" ==== cleanup ====
185 .It Ar cleanup Op Ar filesystem
186 This is a meta-command which executes snapshot, pruning, and reblocking
187 commands on the specified HAMMER filesystem.
188 If no filesystem is specified this command will scan all HAMMER and nullfs
189 mounts, extract PFS id's, and clean-up each PFS found.
191 This command will access a
195 file for each filesystem, creating them if necessary.
196 The default configuration file will create a daily snapshot, do a daily
197 pruning and reblocking run with a fragmentation level of 95%, and
198 a monthly unconditional reblocking run.
200 All operations are limited to 5 minutes per function by default.
201 Reblocking runs are broken down into three separate functions and
203 Also note that this directive will disable snapshots on
210 The defaults may be adjusted by modifying the config file.
211 The pruning and reblocking commands automatically maintain a cyclefile
212 for incremental operation.
213 If you ^C the program the cyclefile will be updated, but a sub-command
214 may continue to run in the background for a few seconds until the HAMMER
215 ioctl detects the interrupt.
217 Work on this command is still in progress.
218 Expected additions: An ability to remove snapshots dynamically as the
219 filesystem becomes full.
221 .It Ar prune Ar softlink-dir
222 Prune the file system based on previously created snapshot softlinks.
223 Pruning is the act of deleting file system history.
227 will delete file system history such that
228 the file system state is retained for the given snapshots,
229 and all history after the latest snapshot,
230 but all other history is deleted.
232 The target directory is expected to contain softlinks pointing to
233 snapshots of the file systems you wish to retain. The directory is scanned
234 non-recursively and the mount points and transaction ids stored in the
235 softlinks are extracted and sorted.
236 The file system is then explicitly pruned according to what is found.
237 Cleaning out portions of the file system is as simple as removing a softlink
242 As a safety measure pruning only occurs if one or more softlinks are found
243 containing the @@ snapshot id extension.
244 Currently the scanned softlink directory must contain softlinks pointing
247 mount. The softlinks may specify absolute or relative
248 paths. Softlinks must use 20-character (@@0x%016llx) transaction ids,
249 as might be returned from
250 .Dq Nm Ar synctid filesystem .
252 Pruning is a per-PFS operation, so a
254 file system and each PFS in it have to be pruned separately.
256 Note that pruning a file system may not immediately free-up space,
257 though typically some space will be freed if a large number of records are
258 pruned out. The file system must be reblocked to completely recover all
261 Example, lets say your snapshot directory contains the following links:
263 lrwxr-xr-x 1 root wheel 29 May 31 17:57 snap1 ->
264 /usr/obj/@@0x10d2cd05b7270d16
266 lrwxr-xr-x 1 root wheel 29 May 31 17:58 snap2 ->
267 /usr/obj/@@0x10d2cd13f3fde98f
269 lrwxr-xr-x 1 root wheel 29 May 31 17:59 snap3 ->
270 /usr/obj/@@0x10d2cd222adee364
273 If you were to run the
275 command on this directory, then the
278 mount will be pruned to retain the above three snapshots.
279 In addition, history for modifications made to the file system older than the oldest
280 snapshot will be destroyed and history for potentially fine-grained modifications made
281 to the file system more recently than the most recent snapshot will be
284 If you then delete the snap2 softlink and rerun the
287 history for modifications pertaining to that snapshot would be destroyed.
288 .\" ==== prune-everything ====
289 .It Ar prune-everything Ar filesystem
290 This command will remove all historical records from the file system.
291 This directive is not normally used on a production system.
292 .\" ==== snapshot ====
293 .It Ar snapshot Ar snapshot-dir
294 .It Ar snapshot Ar filesystem snapshot-dir
295 Takes a snapshot of the file system either explicitly given by
297 or implicitly derived from the
299 argument and creates a symlink in the directory provided by
301 pointing to the snapshot.
304 is not a directory, it is assumed to be a format string
307 with the current time as parameter.
310 refers to an existing directory, a default format string of "snap-%Y%d%m-%H%M"
311 is assumed and used as name for the newly created symlink.
313 Example, assuming that
319 is a file system on its own, the following invocations:
321 hammer snapshot /mysnapshots
323 hammer snapshot /mysnapshots/%Y-%m-%d
325 hammer snapshot /obj /mysnapshots/obj-%Y-%m-%d
328 would create symlinks similar to:
330 /mysnapshots/snap-20080627-1210 -> /@@0x10d2cd05b7270d16
332 /mysnapshots/2008-06-27 -> /@@0x10d2cd05b7270d16
334 /mysnapshots/obj-2008-06-27 -> /obj@@0x10d2cd05b7270d16
336 .\" ==== reblock ====
337 .It Ar reblock Ar filesystem Op Ar fill_percentage
338 .It Ar reblock-btree Ar filesystem Op Ar fill_percentage
339 .It Ar reblock-inodes Ar filesystem Op Ar fill_percentage
340 .It Ar reblock-dirs Ar filesystem Op Ar fill_percentage
341 .It Ar reblock-data Ar filesystem Op Ar fill_percentage
342 Attempt to defragment and free space for reuse by reblocking a live
345 Big blocks cannot be reused by
347 until they are completely free.
348 This command also has the effect of reordering all elements, effectively
349 defragmenting the file system.
351 The default fill percentage is 100% and will cause the file system to be
352 completely defragmented. All specified element types will be reallocated
353 and rewritten. If you wish to quickly free up space instead try specifying
354 a smaller fill percentage, such as 90% or 80% (the
356 suffix is not needed).
358 Since this command may rewrite the entire contents of the disk it is
359 best to do it incrementally from a
365 options to limit the run time.
366 The file system would thus be defragmented over long period of time.
368 It is recommended that separate invocations be used for each data type.
369 B-tree nodes, inodes, and directories are typically the most important
370 elements needing defragmentation. Data can be defragmented over a longer
373 Reblocking is a per-PFS operation, so a
375 file system and each PFS in it have to be reblocked separately.
376 .\" ==== pfs-status ====
377 .It Ar pfs-status Ar dirpath ...
378 Retrieve the mirroring configuration parameters for the specified
380 file systems or pseudo-filesystems.
381 .\" ==== pfs-master ====
382 .It Ar pfs-master Ar dirpath Op options
383 Create a pseudo-filesystem (PFS) inside a
386 Up to 65535 such file systems can be created.
387 Each PFS uses an independent inode numbering space making it suitable
388 for use as a replication source or target.
392 directive creates a PFS that you can read, write, and use as a mirroring
394 .\" ==== pfs-slave ====
395 .It Ar pfs-slave Ar dirpath Op options
396 Create a pseudo-filesystem (PFS) inside a
399 Up to 65535 such file systems can be created.
400 Each PFS uses an independent inode numbering space making it suitable
401 for use as a replication source or target.
405 directive creates a PFS that you can use as a mirroring target.
406 You will not be able to access a slave PFS until you have completed the
407 first mirroring operation with it as the target (its root directory will
408 not exist until then).
410 Access to the pfs-slave via the special softlink,
411 as described in the PFS NOTES below, allows
414 dynamically modify the snapshot transaction id by returning a dynamic result
419 A PFS can only be truly destroyed with the
422 Removing the softlink will not destroy the underlying PFS.
423 .\" ==== pfs-update ====
424 .It Ar pfs-update Ar dirpath Op options
425 Update the configuration parameters for an existing
428 or pseudo-filesystem. Options that may be specified:
429 .Bl -tag -width indent
430 .It sync-beg-tid=0x16llx
431 This is the automatic snapshot access starting transaction id for mirroring slaves.
432 This parameter is normally updated automatically by the
436 It is important to note that accessing a mirroring slave
437 with a transaction id greater than the last fully synchronized transaction
438 id can result in an unreliable snapshot since you will be accessing
439 data that is still undergoing synchronization.
441 Manually modifying this field is dangerous and can result in a broken
443 .It sync-end-tid=0x16llx
444 This is the current synchronization point for mirroring slaves.
445 This parameter is normally updated automatically by the
449 Manually modifying this field is dangerous and can result in a broken
451 .It shared-uuid=<uuid>
452 Set the shared UUID for this file system. All mirrors must have the same
453 shared UUID. For safety purposes the
455 directives will refuse
456 to operate on a target with a different shared UUID.
458 Changing the shared UUID on an existing, non-empty mirroring target,
459 including an empty but not completely pruned target, can lead
460 to corruption of the mirroring target.
461 .It unique-uuid=<uuid>
462 Set the unique UUID for this file system. This UUID should not be used
463 anywhere else, even on exact copies of the file system.
465 Set a descriptive label for this file system.
466 .It snapshots=<string>
467 Specify the snapshots directory which 'hammer cleanup' will use to manage
468 this PFS. The snapshots directory does not need to be configured for
469 PFS masters and will default to "<fs>/snapshots".
471 PFS slaves are mirroring slaves so you cannot configure a snapshots
472 directory on the slave itself to be managed by the slave's machine.
473 In fact, the slave will likely have a 'snapshots' sub-directory mirrored
474 from the master, but that directory contains the configuration the master
475 is using for its copy of the filesystem, not the configuration that we
476 want to use for our slave.
478 It is recommended that "/var/slaves/<name>" be configured for a PFS
479 slave, where <name> is an appropriate label. You can control snapshot
480 retention on your slave independant of the master.
482 Zero out the snapshots directory path for this PFS.
484 .\" ==== pfs-upgrade ====
485 .It Ar pfs-upgrade Ar dirpath
486 Upgrade a PFS from slave to master operation. The PFS will be rolled back
487 to the current end synchronization tid (removing any partial synchronizations),
488 and will then becomes writable.
492 currently supports only single masters and using
493 this command can easily result in file system corruption if you don't
494 know what you are doing.
496 This directive will refuse to run if any programs have open descriptors
497 in the PFS, including programs chdir'd into the PFS.
498 .\" ==== pfs-downgrade ====
499 .It Ar pfs-downgrade Ar dirpath
500 Downgrade a master PFS from master to slave operation. The PFS becomes
501 read-only and access will be locked to its
504 This directive will refuse to run if any programs have open descriptors
505 in the PFS, including programs chdir'd into the PFS.
506 .\" ==== pfs-destroy ====
507 .It Ar pfs-destroy Ar dirpath
508 This permanently destroys a PFS.
510 This directive will refuse to run if any programs have open descriptors
511 in the PFS, including programs chdir'd into the PFS.
512 .\" ==== mirror-read ====
513 .It Ar mirror-read Ar filesystem Op Ar <begin-tid>
514 Generate a mirroring stream to stdout.
515 The stream ends when the transaction id space has been exhausted.
516 .\" ==== mirror-read-stream ====
517 .It Ar mirror-read-stream Ar filesystem Op Ar <begin-tid>
518 Generate a mirroring stream to stdout.
519 Upon completion the stream is paused until new data is synced to the
520 master, then resumed.
521 Operation continues until the pipe is broken.
522 .\" ==== mirror-write ====
523 .It Ar mirror-write Ar filesystem
524 Take a mirroring stream on stdin.
526 This command will fail if the
528 configuration field for the two file systems do not match.
529 .\" ==== mirror-dump ====
536 representation of the mirroring stream.
537 .\" ==== mirror-copy ====
538 .It Ar mirror-copy Ar [[user@]host:]filesystem Ar [[user@]host:]filesystem
539 This is a shortcut which pipes a
543 command. If a remote host specification is made the program forks a
549 on the appropriate host.
550 The source may be a master or slave PFS, and the target must be a slave PFS.
552 This command also established full duplex communication and turns on
553 the two-way protocol feature which automatically negotiates transaction id ranges
554 without having to use a cycle file.
555 If the operation completes successfully the target PFS's
558 be updated. Note that you must re-chdir into the target PFS to see the
559 updated information. If you do not you will still be in the previous snapshot.
560 .\" ==== mirror-stream ====
561 .It Ar mirror-stream Ar [[user@]host:]filesystem Ar [[user@]host:]filesystem
562 This command works similarly to
564 but does not exit unless the pipe is broken.
565 This command will resume the mirroring operation whenever the master is
566 synced. The command is commonly used with
570 options to keep the mirroring target in sync with the source on a continuing
574 .Sh PSEUDO FILESYSTEM (PFS) NOTES
575 The root of a PFS is not hooked into the primary
581 creates a special softlink called "@@PFS%05d" (exactly 10
582 characters long) in the primary
586 then modifies the contents of the softlink as read by
588 and thus what you see with an
590 command or if you were to
593 If the PFS is a master the link reflects the current state of the PFS.
594 If the PFS is a slave the link reflects the last completed snapshot, and the
595 contents of the link will change when the next snapshot is completed, and
598 PFS support is currently very new and experimental. The
601 employs numerous safeties to reduce user foot-shooting.
604 directive requires that the target be configured as a slave and that the
606 field of the mirroring source and target match.
617 utility first appeared in
620 .An Matthew Dillon Aq dillon@backplane.com