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 1.58 2008/11/13 02:04:27 dillon Exp $
40 .Nd HAMMER file system utility
46 .Op Fl f Ar blkdev[:blkdev]*
47 .\" .Op Fl s Ar linkpath
54 This manual page documents the
56 utility which provides miscellaneous functions related to managing a
59 For a general introduction to the
61 file system, its features, and
62 examples on how to set up and maintain one, see
65 The options are as follows:
66 .Bl -tag -width indent
70 Tell the mirror commands to use a 2-way protocol, which allows
71 automatic negotiation of transaction id ranges.
72 This option is automatically enabled by the
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
84 to specify values in kilobytes, megabytes, and gigabytes per second.
85 If no suffix is specified, bytes per second is assumed.
87 When pruning and reblocking you can instruction
89 to start at the object id stored in the specified file.
90 If the file does not exist
92 will start at the beginning.
96 specific period of time and is unable to complete the operation it will
97 write out the current object id so the next run can pick up where it left off.
100 runs to completion it will delete
102 .It Fl f Ar blkdev[:blkdev]*
103 Specify the volumes making up a
107 When maintaining a streaming mirroring this option specifies the
108 minimum delay after a batch ends before the next batch is allowed
110 The default is five seconds.
112 Decrease verboseness.
113 May be specified multiple times.
115 Specify recursion for those commands which support it.
117 When pruning and reblocking you can tell the utility to stop after a
118 certain period of time.
119 This option is used along with the
121 option to prune or reblock a portion of the file system incrementally.
123 Increase verboseness.
124 May be specified multiple times.
126 Force "yes" for any interactive question.
127 .It Fl C Ar cachesize
128 Set the memory cache size for any raw I/O. The default is 16m.
129 A suffix of 'k' for kilobytes and 'm' for megabytes is allowed,
130 else the cache size is specified in bytes.
132 This option is typically only used with diagnostic commands
133 as kernel-supported commands will use the kernel's buffer cache.
136 The commands are as follows:
137 .Bl -tag -width indent
138 .\" ==== synctid ====
139 .It Ar synctid Ar filesystem Op quick
140 Generates a guaranteed, formal 64 bit transaction id representing the
141 current state of the specified
144 The file system will be synced to the media.
148 keyword is specified the file system will be soft-synced, meaning that a
149 crash might still undo the state of the file system as of the transaction
150 id returned but any new modifications will occur after the returned
151 transaction id as expected.
153 .It Ar bstats Op interval
156 B-tree statistics until interrupted.
159 seconds between each display.
160 The default interval is one second.
161 .\" ==== iostats ====
162 .It Ar iostats Op interval
165 I/O statistics until interrupted.
168 seconds between each display.
169 The default interval is one second.
170 .\" ==== history ====
171 .It Ar history Ar path ...
172 Show the modification history for
174 file's inode and data.
178 This command needs the
182 .\" Dump the B-tree, record, large-data, and small-data blockmaps, showing
183 .\" physical block assignments and free space percentages.
184 .\" ==== namekey1 ====
185 .It Ar namekey1 Ar filename
188 64 bit directory hash for the specified file name, using
189 the original directory hash algorithm in version 1 of the filesystem.
190 The low 32 bits are used as an iterator for hash collisions and will be
192 .\" ==== namekey2 ====
193 .It Ar namekey2 Ar filename
196 64 bit directory hash for the specified file name, using
197 the new directory hash algorithm in version 2 of the filesystem.
198 The low 32 bits are still used as an iterator but will start out containing
199 part of the hash key.
200 .\" ==== namekey32 ====
201 .It Ar namekey32 Ar filename
202 Generate the top 32 bits of a
204 64 bit directory hash for the specified file name.
205 .\" ==== cleanup ====
206 .It Ar cleanup Op Ar filesystem ...
207 This is a meta-command which executes snapshot, pruning, and reblocking
208 commands on the specified
213 is specified this command will clean-up all
215 file systems in use, including PFS's.
216 To do this it will scan all
220 mounts, extract PFS id's, and clean-up each PFS found.
222 This command will by default access a
228 creating them if necessary.
229 The format of the configuration file is:
230 .Bd -literal -offset indent
231 snapshots <period> <retention-time> [any]
232 prune <period> <max-runtime>
233 rebalance <period> <max-runtime>
234 reblock <period> <1/3 max-runtime>
235 recopy <period> <1/3 max-runtime>
238 snapshots 1d 60d # 0d 60d for PFS /tmp, /var/tmp, /usr/obj
245 Time is given with a suffix of
251 meaning day, hour, minute and second.
253 If the snapshots directive has a period of 0 and a retention time of 0
254 then snapshot generation is disabled, removal of old snapshots are
255 disabled, and prunes will use
256 .Ar prune-everything .
257 If the snapshots directive has a period of 0 but a non-zero retention time
258 then this command will not create any new snapshots but will remove old
259 snapshots it finds based on the retention time.
261 By default only snapshots in the form: snap-yyyymmdd[-hhmm] are processed.
264 directive is specified as a third argument on the snapshots config line
265 then any softlink of the form *[- or .]yyyymmdd[-hhmm] will be processed.
267 A prune max-runtime of 0 means unlimited.
269 If period hasn't passed since the previous
272 For example a day has passed when midnight is passed (localtime).
280 The default configuration file will create a daily snapshot, do a daily
281 pruning and reblocking run and a monthly recopy run.
282 Reblocking is defragmentation with a level of 95%,
283 and recopy is full defragmentation.
285 By default prune and reblock operations are limited to 5 minutes per function,
286 and recopy operations are limited to 10 minutes per function.
287 Reblocking and recopy runs are each broken down into three separate functions
288 (btree, inodes and data)
289 and are thus by default limited to 15 and 30 minutes respectively.
290 Also note that this directive will by default disable snapshots on
297 The defaults may be adjusted by modifying the
300 The pruning and reblocking commands automatically maintain a cyclefile
301 for incremental operation.
302 If you interrupt (^C) the program the cyclefile will be updated, but a sub-command
303 may continue to run in the background for a few seconds until the
305 ioctl detects the interrupt.
308 PFS option can be set to use another location for the snapshots directory.
310 Work on this command is still in progress.
311 Expected additions: An ability to remove snapshots dynamically as the
312 file system becomes full.
313 .\" ==== snapshot ====
314 .It Ar snapshot Oo Ar filesystem Oc Ar snapshot-dir
315 Takes a snapshot of the file system either explicitly given by
317 or implicitly derived from the
319 argument and creates a symlink in the directory provided by
321 pointing to the snapshot.
324 is not a directory, it is assumed to be a format string passed to
326 with the current time as parameter.
329 refers to an existing directory, a default format string of "snap-%Y%d%m-%H%M"
330 is assumed and used as name for the newly created symlink.
332 Snapshot is a per PFS operation, so a
334 file system and each PFS in it have to be snapshot separately.
336 Example, assuming that
342 is a file system on its own, the following invocations:
343 .Bd -literal -offset indent
344 hammer snapshot /mysnapshots
346 hammer snapshot /mysnapshots/%Y-%m-%d
348 hammer snapshot /obj /mysnapshots/obj-%Y-%m-%d
351 would create symlinks similar to:
352 .Bd -literal -offset indent
353 /mysnapshots/snap-20080627-1210 -> /@@0x10d2cd05b7270d16
355 /mysnapshots/2008-06-27 -> /@@0x10d2cd05b7270d16
357 /mysnapshots/obj-2008-06-27 -> /obj@@0x10d2cd05b7270d16
360 .It Ar prune Ar softlink-dir
361 Prune the file system based on previously created snapshot softlinks.
362 Pruning is the act of deleting file system history.
366 will delete file system history such that
367 the file system state is retained for the given snapshots,
368 and all history after the latest snapshot,
369 but all other history is deleted.
371 The target directory is expected to contain softlinks pointing to
372 snapshots of the file systems you wish to retain.
373 The directory is scanned non-recursively and the mount points and
374 transaction ids stored in the softlinks are extracted and sorted.
375 The file system is then explicitly pruned according to what is found.
376 Cleaning out portions of the file system is as simple as removing a softlink
381 As a safety measure pruning only occurs if one or more softlinks are found
382 containing the @@ snapshot id extension.
383 Currently the scanned softlink directory must contain softlinks pointing
387 The softlinks may specify absolute or relative paths.
388 Softlinks must use 20-character (@@0x%016llx) transaction ids,
389 as might be returned from
390 .Dq Nm Ar synctid filesystem .
392 Pruning is a per PFS operation, so a
394 file system and each PFS in it have to be pruned separately.
396 Note that pruning a file system may not immediately free-up space,
397 though typically some space will be freed if a large number of records are
399 The file system must be reblocked to completely recover all available space.
401 Example, lets say your snapshot directory contains the following links:
402 .Bd -literal -offset indent
403 lrwxr-xr-x 1 root wheel 29 May 31 17:57 snap1 ->
404 /usr/obj/@@0x10d2cd05b7270d16
406 lrwxr-xr-x 1 root wheel 29 May 31 17:58 snap2 ->
407 /usr/obj/@@0x10d2cd13f3fde98f
409 lrwxr-xr-x 1 root wheel 29 May 31 17:59 snap3 ->
410 /usr/obj/@@0x10d2cd222adee364
413 If you were to run the
415 command on this directory, then the
418 mount will be pruned to retain the above three snapshots.
419 In addition, history for modifications made to the file system older than
420 the oldest snapshot will be destroyed and history for potentially fine-grained
421 modifications made to the file system more recently than the most recent
422 snapshot will be retained.
424 If you then delete the
426 softlink and rerun the
429 history for modifications pertaining to that snapshot would be destroyed.
430 .\" ==== prune-everything ====
431 .It Ar prune-everything Ar filesystem
432 This command will remove all historical records from the file system.
433 This directive is not normally used on a production system.
434 .\" ==== rebalance ====
435 .It Ar rebalance Ar filesystem Op Ar saturation_level
436 This command will rebalance the B-Tree, nodes with small numbers of
437 elements will be combined and element counts will be smoothed out
440 The saturation level is a percentage between 50 and 100. The default
442 .\" ==== reblock ====
443 .It Ar reblock Ar filesystem Op Ar fill_percentage
444 .It Ar reblock-btree Ar filesystem Op Ar fill_percentage
445 .It Ar reblock-inodes Ar filesystem Op Ar fill_percentage
446 .It Ar reblock-dirs Ar filesystem Op Ar fill_percentage
447 .It Ar reblock-data Ar filesystem Op Ar fill_percentage
448 Attempt to defragment and free space for reuse by reblocking a live
451 Big blocks cannot be reused by
453 until they are completely free.
454 This command also has the effect of reordering all elements, effectively
455 defragmenting the file system.
457 The default fill percentage is 100% and will cause the file system to be
458 completely defragmented.
459 All specified element types will be reallocated and rewritten.
460 If you wish to quickly free up space instead try specifying
461 a smaller fill percentage, such as 90% or 80% (the
463 suffix is not needed).
465 Since this command may rewrite the entire contents of the disk it is
466 best to do it incrementally from a
472 options to limit the run time.
473 The file system would thus be defragmented over long period of time.
475 It is recommended that separate invocations be used for each data type.
476 B-tree nodes, inodes, and directories are typically the most important
477 elements needing defragmentation.
478 Data can be defragmented over a longer period of time.
480 Reblocking is a per PFS operation, so a
482 file system and each PFS in it have to be reblocked separately.
483 .\" ==== pfs-status ====
484 .It Ar pfs-status Ar dirpath ...
485 Retrieve the mirroring configuration parameters for the specified
487 file systems or pseudo-filesystems (PFS's).
488 .\" ==== pfs-master ====
489 .It Ar pfs-master Ar dirpath Op options
490 Create a pseudo-filesystem (PFS) inside a
493 Up to 65535 such file systems can be created.
494 Each PFS uses an independent inode numbering space making it suitable
495 for use as a replication source or target.
499 directive creates a PFS that you can read, write, and use as a mirroring
502 It is recommended to use a
504 mount to access a PFS, for more information see
506 .\" ==== pfs-slave ====
507 .It Ar pfs-slave Ar dirpath Op options
508 Create a pseudo-filesystem (PFS) inside a
511 Up to 65535 such file systems can be created.
512 Each PFS uses an independent inode numbering space making it suitable
513 for use as a replication source or target.
517 directive creates a PFS that you can use as a mirroring target.
518 You will not be able to access a slave PFS until you have completed the
519 first mirroring operation with it as the target (its root directory will
520 not exist until then).
522 Access to the pfs-slave via the special softlink,
528 dynamically modify the snapshot transaction id by returning a dynamic result
533 A PFS can only be truly destroyed with the
536 Removing the softlink will not destroy the underlying PFS.
538 It is recommended to use a
540 mount to access a PFS, for more information see
542 .\" ==== pfs-update ====
543 .It Ar pfs-update Ar dirpath Op options
544 Update the configuration parameters for an existing
547 or pseudo-filesystem.
548 Options that may be specified:
549 .Bl -tag -width indent
550 .It sync-beg-tid=0x16llx
551 This is the automatic snapshot access starting transaction id for
553 This parameter is normally updated automatically by the
557 It is important to note that accessing a mirroring slave
558 with a transaction id greater than the last fully synchronized transaction
559 id can result in an unreliable snapshot since you will be accessing
560 data that is still undergoing synchronization.
562 Manually modifying this field is dangerous and can result in a broken
564 .It sync-end-tid=0x16llx
565 This is the current synchronization point for mirroring slaves.
566 This parameter is normally updated automatically by the
570 Manually modifying this field is dangerous and can result in a broken mirror.
571 .It shared-uuid=<uuid>
572 Set the shared UUID for this file system.
573 All mirrors must have the same shared UUID.
574 For safety purposes the
576 directives will refuse to operate on a target with a different shared UUID.
578 Changing the shared UUID on an existing, non-empty mirroring target,
579 including an empty but not completely pruned target,
580 can lead to corruption of the mirroring target.
581 .It unique-uuid=<uuid>
582 Set the unique UUID for this file system.
583 This UUID should not be used anywhere else,
584 even on exact copies of the file system.
586 Set a descriptive label for this file system.
587 .It snapshots=<string>
588 Specify the snapshots directory which
591 will use to manage this PFS.
592 The snapshots directory does not need to be configured for
593 PFS masters and will default to
594 .Pa <pfs>/snapshots .
596 PFS slaves are mirroring slaves so you cannot configure a snapshots
597 directory on the slave itself to be managed by the slave's machine.
598 In fact, the slave will likely have a
600 sub-directory mirrored
601 from the master, but that directory contains the configuration the master
602 is using for its copy of the file system, not the configuration that we
603 want to use for our slave.
605 It is recommended that
606 .Pa <fs>/var/slaves/<name>
607 be configured for a PFS slave, where
613 is an appropriate label.
614 You can control snapshot retention on your slave independent of the master.
616 Zero out the snapshots directory path for this PFS.
618 .\" ==== pfs-upgrade ====
619 .It Ar pfs-upgrade Ar dirpath
620 Upgrade a PFS from slave to master operation.
621 The PFS will be rolled back to the current end synchronization tid
622 (removing any partial synchronizations), and will then become writable.
626 currently supports only single masters and using
627 this command can easily result in file system corruption
628 if you don't know what you are doing.
630 This directive will refuse to run if any programs have open descriptors
631 in the PFS, including programs chdir'd into the PFS.
632 .\" ==== pfs-downgrade ====
633 .It Ar pfs-downgrade Ar dirpath
634 Downgrade a master PFS from master to slave operation
635 The PFS becomes read-only and access will be locked to its
638 This directive will refuse to run if any programs have open descriptors
639 in the PFS, including programs chdir'd into the PFS.
640 .\" ==== pfs-destroy ====
641 .It Ar pfs-destroy Ar dirpath
642 This permanently destroys a PFS.
644 This directive will refuse to run if any programs have open descriptors
645 in the PFS, including programs chdir'd into the PFS.
646 .\" ==== mirror-read ====
647 .It Ar mirror-read Ar filesystem Op Ar <begin-tid>
648 Generate a mirroring stream to stdout.
649 The stream ends when the transaction id space has been exhausted.
650 .\" ==== mirror-read-stream ====
651 .It Ar mirror-read-stream Ar filesystem Op Ar <begin-tid>
652 Generate a mirroring stream to stdout.
653 Upon completion the stream is paused until new data is synced to the
654 master, then resumed.
655 Operation continues until the pipe is broken.
656 .\" ==== mirror-write ====
657 .It Ar mirror-write Ar filesystem
658 Take a mirroring stream on stdin.
660 This command will fail if the
662 configuration field for the two file systems do not match.
664 If the target PFS does not exist this command will ask you whether
665 you want to create a compatible PFS slave for the target or not.
666 .\" ==== mirror-dump ====
672 to dump an ASCII representation of the mirroring stream.
673 .\" ==== mirror-copy ====
674 .It Ar mirror-copy Ar [[user@]host:]filesystem Ar [[user@]host:]filesystem
675 This is a shortcut which pipes a
680 If a remote host specification is made the program forks a
686 on the appropriate host.
687 The source may be a master or slave PFS, and the target must be a slave PFS.
689 This command also established full duplex communication and turns on
690 the two-way protocol feature which automatically negotiates transaction id
691 ranges without having to use a cyclefile.
692 If the operation completes successfully the target PFS's
695 Note that you must re-chdir into the target PFS to see the updated information.
696 If you do not you will still be in the previous snapshot.
698 If the target PFS does not exist this command will ask you whether
699 you want to create a compatible PFS slave for the target or not.
700 .\" ==== mirror-stream ====
701 .It Ar mirror-stream Ar [[user@]host:]filesystem Ar [[user@]host:]filesystem
702 This command works similarly to
704 but does not exit unless the pipe is broken.
705 This command will resume the mirroring operation whenever the master is synced.
706 The command is commonly used with
710 options to keep the mirroring target in sync with the source on a continuing
712 .\" ==== version ====
713 .It Ar version Ar filesystem
714 This command returns the
716 filesystem version for the specified
717 filesystem as well as the range of versions supported in the kernel.
720 option may be used to remove the summary at the end.
721 .\" ==== version-upgrade ====
722 .It Ar version-upgrade Ar filesystem Ar version Op Ar force
723 This command upgrades the
725 filesystem to the specified version.
726 Once upgraded a filesystem may not be downgraded.
727 If you wish to upgrade a filesystem to a version greater or equal to the
728 work-in-progress version number you must specify the
731 Use of WIP versions should be relegated to testing and may require wiping
732 the filesystem as development progresses, even though the WIP version might
735 NOTE! This command operates on the entire
737 filesystem and is not a per-PFS operation.
738 All PFS's will be affected.
739 .Bl -tag -width indent
742 default version, first
746 Work-in-progress version.
747 This version is developing a new directory hash key.
751 .Sh PSEUDO FILESYSTEM (PFS) NOTES
752 The root of a PFS is not hooked into the primary
754 file system as a directory.
757 creates a special softlink called "@@PFS%05d" (exactly 10 characters long)
762 then modifies the contents of the softlink as read by
764 and thus what you see with an
766 command or if you were to
769 If the PFS is a master the link reflects the current state of the PFS.
770 If the PFS is a slave the link reflects the last completed snapshot, and the
771 contents of the link will change when the next snapshot is completed, and
774 PFS support is currently very new and experimental.
777 utility employs numerous safeties to reduce user foot-shooting.
780 directive requires that the target be configured as a slave and that the
782 field of the mirroring source and target match.
784 .Bl -tag -width ".It Pa <fs>/var/slaves/<name>" -compact
786 default per PFS snapshots directory
787 .It Pa <snapshots>/config
791 .It Pa <fs>/var/slaves/<name>
792 recommended slave PFS snapshots directory
799 .Xr periodic.conf 5 ,
806 utility first appeared in
809 .An Matthew Dillon Aq dillon@backplane.com
projects / dragonfly.git / blob