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 1.55 2008/10/08 21:01:54 thomas 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.
70 This option is automatically enabled by the
74 Specify a bandwidth limit in bytes per second for mirroring streams.
75 This option is typically used to prevent batch mirroring operations from
76 loading down the machine.
77 The bandwidth may be suffixed with
82 to specify values in kilobytes, megabytes, and gigabytes per second.
84 When pruning and reblocking you can instruction
86 to start at the object id stored in the specified file.
87 If the file does not exist
89 will start at the beginning.
93 specific period of time and is unable to complete the operation it will
94 write out the current object id so the next run can pick up where it left off.
97 runs to completion it will delete
99 .It Fl f Ar blkdev[:blkdev]*
100 Specify the volumes making up a
104 When maintaining a streaming mirroring this option specifies the
105 minimum delay after a batch ends before the next batch is allowed
107 The default is five seconds.
109 Decrease verbosement.
110 May be specified multiple times.
112 Specify recursion for those commands which support it.
114 When pruning and reblocking you can tell the utility to stop after a
115 certain period of time.
116 This option is used along with the
118 option to prune or reblock a portion of the file system incrementally.
120 Increase verboseness.
121 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
132 The file system will 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.
166 This command needs the
170 .\" Dump the B-tree, record, large-data, and small-data blockmaps, showing
171 .\" physical block assignments and free space percentages.
172 .\" ==== namekey ====
173 .It Ar namekey Ar filename
176 64 bit directory hash for the specified file name.
177 The low 32 bits are used as an iterator for hash collisions and will be
179 .\" ==== namekey32 ====
180 .It Ar namekey32 Ar filename
181 Generate the top 32 bits of a
183 64 bit directory hash for the specified file name.
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
192 is specified this command will clean-up all
194 file systems in use, including PFS's.
195 To do this it will scan all
199 mounts, extract PFS id's, and clean-up each PFS found.
201 This command will by default access a
207 creating them if necessary.
208 The format of the configuration file is:
209 .Bd -literal -offset indent
210 snapshots <period> <retensiontime>
211 prune <period> <max-runtime>
212 reblock <period> <1/3 max-runtime>
213 recopy <period> <1/3 max-runtime>
216 snapshots 1d 60d # 0d 60d for PFS /tmp, /var/tmp, /usr/obj
222 Time is given with a suffix of
228 meaning day, hour, minute and second.
229 A snapshots period of 0 disables snapshot generation and prunes using
230 .Ar prune-everything ,
231 if no snapshots exists.
232 A prune max-runtime of 0 means unlimited.
234 If period hasn't passed since previous
237 For example a day has passed when midnight is passed (localtime).
238 It is recommended to run
240 once a day, depending on the configured period, from a
244 .Bd -literal -offset indent
245 15 2 * * * hammer cleanup >>/var/log/hammer.log 2>&1
248 The default configuration file will create a daily snapshot, do a daily
249 pruning and reblocking run and a monthly recopy run.
250 Reblocking is defragmentation with a level of 95%,
251 and recopy is full defragmentation.
253 By default prune and reblock operations are limited to 5 minutes per function,
254 and recopy operations are limited to 10 minutes per function.
255 Reblocking and recopy runs are each broken down into three separate functions
256 (btree, inodes and data)
257 and are thus by default limited to 15 and 30 minutes respectively.
258 Also note that this directive will by default disable snapshots on
265 The defaults may be adjusted by modifying the
268 The pruning and reblocking commands automatically maintain a cyclefile
269 for incremental operation.
270 If you interrupt (^C) the program the cyclefile will be updated, but a sub-command
271 may continue to run in the background for a few seconds until the
273 ioctl detects the interrupt.
276 PFS option can be set to use another location for the snapshots directory.
278 Work on this command is still in progress.
279 Expected additions: An ability to remove snapshots dynamically as the
280 file system becomes full.
281 .\" ==== snapshot ====
282 .It Ar snapshot Oo Ar filesystem Oc Ar snapshot-dir
283 Takes a snapshot of the file system either explicitly given by
285 or implicitly derived from the
287 argument and creates a symlink in the directory provided by
289 pointing to the snapshot.
292 is not a directory, it is assumed to be a format string passed to
294 with the current time as parameter.
297 refers to an existing directory, a default format string of "snap-%Y%d%m-%H%M"
298 is assumed and used as name for the newly created symlink.
300 Snapshot is a per PFS operation, so a
302 file system and each PFS in it have to be snapshot separately.
304 Example, assuming that
310 is a file system on its own, the following invocations:
311 .Bd -literal -offset indent
312 hammer snapshot /mysnapshots
314 hammer snapshot /mysnapshots/%Y-%m-%d
316 hammer snapshot /obj /mysnapshots/obj-%Y-%m-%d
319 would create symlinks similar to:
320 .Bd -literal -offset indent
321 /mysnapshots/snap-20080627-1210 -> /@@0x10d2cd05b7270d16
323 /mysnapshots/2008-06-27 -> /@@0x10d2cd05b7270d16
325 /mysnapshots/obj-2008-06-27 -> /obj@@0x10d2cd05b7270d16
328 .It Ar prune Ar softlink-dir
329 Prune the file system based on previously created snapshot softlinks.
330 Pruning is the act of deleting file system history.
334 will delete file system history such that
335 the file system state is retained for the given snapshots,
336 and all history after the latest snapshot,
337 but all other history is deleted.
339 The target directory is expected to contain softlinks pointing to
340 snapshots of the file systems you wish to retain.
341 The directory is scanned non-recursively and the mount points and
342 transaction ids stored in the softlinks are extracted and sorted.
343 The file system is then explicitly pruned according to what is found.
344 Cleaning out portions of the file system is as simple as removing a softlink
349 As a safety measure pruning only occurs if one or more softlinks are found
350 containing the @@ snapshot id extension.
351 Currently the scanned softlink directory must contain softlinks pointing
355 The softlinks may specify absolute or relative paths.
356 Softlinks must use 20-character (@@0x%016llx) transaction ids,
357 as might be returned from
358 .Dq Nm Ar synctid filesystem .
360 Pruning is a per PFS operation, so a
362 file system and each PFS in it have to be pruned separately.
364 Note that pruning a file system may not immediately free-up space,
365 though typically some space will be freed if a large number of records are
367 The file system must be reblocked to completely recover all available space.
369 Example, lets say your snapshot directory contains the following links:
370 .Bd -literal -offset indent
371 lrwxr-xr-x 1 root wheel 29 May 31 17:57 snap1 ->
372 /usr/obj/@@0x10d2cd05b7270d16
374 lrwxr-xr-x 1 root wheel 29 May 31 17:58 snap2 ->
375 /usr/obj/@@0x10d2cd13f3fde98f
377 lrwxr-xr-x 1 root wheel 29 May 31 17:59 snap3 ->
378 /usr/obj/@@0x10d2cd222adee364
381 If you were to run the
383 command on this directory, then the
386 mount will be pruned to retain the above three snapshots.
387 In addition, history for modifications made to the file system older than
388 the oldest snapshot will be destroyed and history for potentially fine-grained
389 modifications made to the file system more recently than the most recent
390 snapshot will be retained.
392 If you then delete the
394 softlink and rerun the
397 history for modifications pertaining to that snapshot would be destroyed.
398 .\" ==== prune-everything ====
399 .It Ar prune-everything Ar filesystem
400 This command will remove all historical records from the file system.
401 This directive is not normally used on a production system.
402 .\" ==== reblock ====
403 .It Ar reblock Ar filesystem Op Ar fill_percentage
404 .It Ar reblock-btree Ar filesystem Op Ar fill_percentage
405 .It Ar reblock-inodes Ar filesystem Op Ar fill_percentage
406 .It Ar reblock-dirs Ar filesystem Op Ar fill_percentage
407 .It Ar reblock-data Ar filesystem Op Ar fill_percentage
408 Attempt to defragment and free space for reuse by reblocking a live
411 Big blocks cannot be reused by
413 until they are completely free.
414 This command also has the effect of reordering all elements, effectively
415 defragmenting the file system.
417 The default fill percentage is 100% and will cause the file system to be
418 completely defragmented.
419 All specified element types will be reallocated and rewritten.
420 If you wish to quickly free up space instead try specifying
421 a smaller fill percentage, such as 90% or 80% (the
423 suffix is not needed).
425 Since this command may rewrite the entire contents of the disk it is
426 best to do it incrementally from a
432 options to limit the run time.
433 The file system would thus be defragmented over long period of time.
435 It is recommended that separate invocations be used for each data type.
436 B-tree nodes, inodes, and directories are typically the most important
437 elements needing defragmentation.
438 Data can be defragmented over a longer period of time.
440 Reblocking is a per PFS operation, so a
442 file system and each PFS in it have to be reblocked separately.
443 .\" ==== pfs-status ====
444 .It Ar pfs-status Ar dirpath ...
445 Retrieve the mirroring configuration parameters for the specified
447 file systems or pseudo-filesystems (PFS's).
448 .\" ==== pfs-master ====
449 .It Ar pfs-master Ar dirpath Op options
450 Create a pseudo-filesystem (PFS) inside a
453 Up to 65535 such file systems can be created.
454 Each PFS uses an independent inode numbering space making it suitable
455 for use as a replication source or target.
459 directive creates a PFS that you can read, write, and use as a mirroring
462 It is recommended to use a
464 mount to access a PFS, for more information see
466 .\" ==== pfs-slave ====
467 .It Ar pfs-slave Ar dirpath Op options
468 Create a pseudo-filesystem (PFS) inside a
471 Up to 65535 such file systems can be created.
472 Each PFS uses an independent inode numbering space making it suitable
473 for use as a replication source or target.
477 directive creates a PFS that you can use as a mirroring target.
478 You will not be able to access a slave PFS until you have completed the
479 first mirroring operation with it as the target (its root directory will
480 not exist until then).
482 Access to the pfs-slave via the special softlink,
488 dynamically modify the snapshot transaction id by returning a dynamic result
493 A PFS can only be truly destroyed with the
496 Removing the softlink will not destroy the underlying PFS.
498 It is recommended to use a
500 mount to access a PFS, for more information see
502 .\" ==== pfs-update ====
503 .It Ar pfs-update Ar dirpath Op options
504 Update the configuration parameters for an existing
507 or pseudo-filesystem.
508 Options that may be specified:
509 .Bl -tag -width indent
510 .It sync-beg-tid=0x16llx
511 This is the automatic snapshot access starting transaction id for
513 This parameter is normally updated automatically by the
517 It is important to note that accessing a mirroring slave
518 with a transaction id greater than the last fully synchronized transaction
519 id can result in an unreliable snapshot since you will be accessing
520 data that is still undergoing synchronization.
522 Manually modifying this field is dangerous and can result in a broken
524 .It sync-end-tid=0x16llx
525 This is the current synchronization point for mirroring slaves.
526 This parameter is normally updated automatically by the
530 Manually modifying this field is dangerous and can result in a broken mirror.
531 .It shared-uuid=<uuid>
532 Set the shared UUID for this file system.
533 All mirrors must have the same shared UUID.
534 For safety purposes the
536 directives will refuse to operate on a target with a different shared UUID.
538 Changing the shared UUID on an existing, non-empty mirroring target,
539 including an empty but not completely pruned target,
540 can lead to corruption of the mirroring target.
541 .It unique-uuid=<uuid>
542 Set the unique UUID for this file system.
543 This UUID should not be used anywhere else,
544 even on exact copies of the file system.
546 Set a descriptive label for this file system.
547 .It snapshots=<string>
548 Specify the snapshots directory which
551 will use to manage this PFS.
552 The snapshots directory does not need to be configured for
553 PFS masters and will default to
554 .Pa <pfs>/snapshots .
556 PFS slaves are mirroring slaves so you cannot configure a snapshots
557 directory on the slave itself to be managed by the slave's machine.
558 In fact, the slave will likely have a
560 sub-directory mirrored
561 from the master, but that directory contains the configuration the master
562 is using for its copy of the file system, not the configuration that we
563 want to use for our slave.
565 It is recommended that
566 .Pa <fs>/var/slaves/<name>
567 be configured for a PFS slave, where
573 is an appropriate label.
574 You can control snapshot retention on your slave independent of the master.
576 Zero out the snapshots directory path for this PFS.
578 .\" ==== pfs-upgrade ====
579 .It Ar pfs-upgrade Ar dirpath
580 Upgrade a PFS from slave to master operation.
581 The PFS will be rolled back to the current end synchronization tid
582 (removing any partial synchronizations), and will then become writable.
586 currently supports only single masters and using
587 this command can easily result in file system corruption
588 if you don't know what you are doing.
590 This directive will refuse to run if any programs have open descriptors
591 in the PFS, including programs chdir'd into the PFS.
592 .\" ==== pfs-downgrade ====
593 .It Ar pfs-downgrade Ar dirpath
594 Downgrade a master PFS from master to slave operation
595 The PFS becomes read-only and access will be locked to its
598 This directive will refuse to run if any programs have open descriptors
599 in the PFS, including programs chdir'd into the PFS.
600 .\" ==== pfs-destroy ====
601 .It Ar pfs-destroy Ar dirpath
602 This permanently destroys a PFS.
604 This directive will refuse to run if any programs have open descriptors
605 in the PFS, including programs chdir'd into the PFS.
606 .\" ==== mirror-read ====
607 .It Ar mirror-read Ar filesystem Op Ar <begin-tid>
608 Generate a mirroring stream to stdout.
609 The stream ends when the transaction id space has been exhausted.
610 .\" ==== mirror-read-stream ====
611 .It Ar mirror-read-stream Ar filesystem Op Ar <begin-tid>
612 Generate a mirroring stream to stdout.
613 Upon completion the stream is paused until new data is synced to the
614 master, then resumed.
615 Operation continues until the pipe is broken.
616 .\" ==== mirror-write ====
617 .It Ar mirror-write Ar filesystem
618 Take a mirroring stream on stdin.
620 This command will fail if the
622 configuration field for the two file systems do not match.
623 .\" ==== mirror-dump ====
629 to dump an ASCII representation of the mirroring stream.
630 .\" ==== mirror-copy ====
631 .It Ar mirror-copy Ar [[user@]host:]filesystem Ar [[user@]host:]filesystem
632 This is a shortcut which pipes a
637 If a remote host specification is made the program forks a
643 on the appropriate host.
644 The source may be a master or slave PFS, and the target must be a slave PFS.
646 This command also established full duplex communication and turns on
647 the two-way protocol feature which automatically negotiates transaction id
648 ranges without having to use a cyclefile.
649 If the operation completes successfully the target PFS's
652 Note that you must re-chdir into the target PFS to see the updated information.
653 If you do not you will still be in the previous snapshot.
654 .\" ==== mirror-stream ====
655 .It Ar mirror-stream Ar [[user@]host:]filesystem Ar [[user@]host:]filesystem
656 This command works similarly to
658 but does not exit unless the pipe is broken.
659 This command will resume the mirroring operation whenever the master is synced.
660 The command is commonly used with
664 options to keep the mirroring target in sync with the source on a continuing
668 .Sh PSEUDO FILESYSTEM (PFS) NOTES
669 The root of a PFS is not hooked into the primary
671 file system as a directory.
674 creates a special softlink called "@@PFS%05d" (exactly 10 characters long)
679 then modifies the contents of the softlink as read by
681 and thus what you see with an
683 command or if you were to
686 If the PFS is a master the link reflects the current state of the PFS.
687 If the PFS is a slave the link reflects the last completed snapshot, and the
688 contents of the link will change when the next snapshot is completed, and
691 PFS support is currently very new and experimental.
694 utility employs numerous safeties to reduce user foot-shooting.
697 directive requires that the target be configured as a slave and that the
699 field of the mirroring source and target match.
703 .Bl -tag -width ".It Pa <fs>/var/slaves/<name>" -compact
705 default per PFS snapshots directory
706 .It Pa <snapshots>/config
710 .It Pa <fs>/var/slaves/<name>
711 recommended slave PFS snapshots directory
722 utility first appeared in
725 .An Matthew Dillon Aq dillon@backplane.com