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
47 .Op Fl C Ar cachesize Ns Op Ns Cm \&: Ns Ar readahead
49 .Op Fl f Ar blkdev Ns Oo Ns Cm \&: Ns Ar blkdev Oc Ns *
50 .\" .Op Fl s Ar linkpath
56 This manual page documents the
58 utility which provides miscellaneous functions related to managing a
61 For a general introduction to the
63 file system, its features, and
64 examples on how to set up and maintain one, see
67 The options are as follows:
68 .Bl -tag -width indent
72 Tell the mirror commands to use a 2-way protocol, which allows
73 automatic negotiation of transaction id ranges.
74 This option is automatically enabled by the
80 will not attempt to break-up large initial bulk transfers into smaller pieces.
81 This can save time but if the link is lost in the middle of the
82 initial bulk transfer you will have to start over from scratch.
84 Specify a bandwidth limit in bytes per second for mirroring streams.
85 This option is typically used to prevent batch mirroring operations from
86 loading down the machine.
87 The bandwidth may be suffixed with
91 to specify values in kilobytes, megabytes, and gigabytes per second.
92 If no suffix is specified, bytes per second is assumed.
93 .It Fl C Ar cachesize Ns Op Ns Cm \&: Ns Ar readahead
94 Set the memory cache size for any raw
101 for megabytes is allowed,
102 else the cache size is specified in bytes.
104 The read-behind/read-ahead defaults to 4
108 This option is typically only used with diagnostic commands
109 as kernel-supported commands will use the kernel's buffer cache.
110 .It Fl c Ar cyclefile
111 When pruning and reblocking you can instruction
113 to start at the object id stored in the specified file.
114 If the file does not exist
116 will start at the beginning.
120 specific period of time and is unable to complete the operation it will
121 write out the current object id so the next run can pick up where it left off.
124 runs to completion it will delete
126 .It Fl f Ar blkdev Ns Oo Ns Cm \&: Ns Ar blkdev Oc Ns *
127 Specify the volumes making up a
131 When maintaining a streaming mirroring this option specifies the
132 minimum delay after a batch ends before the next batch is allowed
134 The default is five seconds.
136 Decrease verboseness.
137 May be specified multiple times.
139 Specify recursion for those commands which support it.
141 When pruning and reblocking you can tell the utility to stop after a
142 certain period of time.
143 This option is used along with the
145 option to prune or reblock a portion of the file system incrementally.
147 Increase verboseness.
148 May be specified multiple times.
150 Force "yes" for any interactive question.
153 The commands are as follows:
154 .Bl -tag -width indent
155 .\" ==== synctid ====
156 .It Cm synctid Ar filesystem Op Cm quick
157 Generates a guaranteed, formal 64 bit transaction id representing the
158 current state of the specified
161 The file system will be synced to the media.
165 keyword is specified the file system will be soft-synced, meaning that a
166 crash might still undo the state of the file system as of the transaction
167 id returned but any new modifications will occur after the returned
168 transaction id as expected.
170 This operation does not create a snapshot.
171 It is meant to be used
172 to track temporary fine-grained changes to a subset of files and
173 will only remain valid for
175 snapshot access purposes for the
177 period configured for the PFS.
178 If you desire a real snapshot then the
180 directive may be what you are looking for.
182 .It Cm bstats Op Ar interval
185 B-tree statistics until interrupted.
188 seconds between each display.
189 The default interval is one second.
190 .\" ==== iostats ====
191 .It Cm iostats Op Ar interval
195 statistics until interrupted.
198 seconds between each display.
199 The default interval is one second.
200 .\" ==== history ====
201 .It Cm history Ar path ...
202 Show the modification history for
204 file's inode and data.
205 .\" ==== blockmap ====
207 Dump the blockmap for the file system.
210 blockmap is two-layer
211 blockmap representing the maximum possible file system size of 1 Exabyte.
212 Needless to say the second layer is only present for blocks which exist.
214 blockmap represents 8-Megabyte blocks, called big-blocks.
215 Each big-block has an append
216 point, a free byte count, and a typed zone id which allows content to be
217 reverse engineered to some degree.
221 allocations essentially appended to a selected big-block using
222 the append offset and deducted from the free byte count.
223 When space is freed the free byte count is adjusted but
225 does not track holes in big-blocks for reallocation.
226 A big-block must be completely freed, either
227 through normal file system operations or through reblocking, before
230 Data blocks can be shared by deducting the space used from the free byte
231 count for each shared references, though
233 does not yet make use of this feature.
234 This means the free byte count can legally go negative.
236 This command needs the
240 .It Cm show Op Ar lo Ns Cm \&: Ns Ar objid
242 By default this command will validate all B-Tree
243 linkages and CRCs, including data CRCs, and will report the most verbose
244 information it can dig up.
245 Any errors will show up with a
247 in column 1 along with various
250 If you specify a localization and object id field the dump will
251 search for the key printing nodes as it recurses down, and then
252 will iterate forwards.
256 the command will report less information about the inode contents.
260 the command will not report the content of the inode or other typed
265 the command will not report volume header information, big-block fill
266 ratios, mirror transaction ids, or report or check data CRCs.
267 B-tree CRCs and linkages are still checked.
269 This command needs the
273 .\" Dump the B-tree, record, large-data, and small-data blockmaps, showing
274 .\" physical block assignments and free space percentages.
275 .\" ==== namekey1 ====
276 .It Cm namekey1 Ar filename
279 64 bit directory hash for the specified file name, using
280 the original directory hash algorithm in version 1 of the file system.
281 The low 32 bits are used as an iterator for hash collisions and will be
283 .\" ==== namekey2 ====
284 .It Cm namekey2 Ar filename
287 64 bit directory hash for the specified file name, using
288 the new directory hash algorithm in version 2 of the file system.
289 The low 32 bits are still used as an iterator but will start out containing
290 part of the hash key.
291 .\" ==== namekey32 ====
292 .It Cm namekey32 Ar filename
293 Generate the top 32 bits of a
295 64 bit directory hash for the specified file name.
298 Shows extended information about all the mounted
301 At the moment volume identification, big-blocks information, space and
302 and pseudo-filesystem (PFS) details are shown.
303 .\" ==== cleanup ====
304 .It Cm cleanup Op Ar filesystem ...
305 This is a meta-command which executes snapshot, prune, rebalance and reblock
306 commands on the specified
311 is specified this command will clean-up all
313 file systems in use, including PFS's.
314 To do this it will scan all
318 mounts, extract PFS id's, and clean-up each PFS found.
320 This command will access a snapshots
321 directory and a configuration file for each
323 creating them if necessary.
324 .Bl -tag -width indent
325 .It Nm HAMMER No version 2-
326 The configuration file is
328 in the snapshots directory which defaults to
329 .Pa <pfs>/snapshots .
330 .It Nm HAMMER No version 3+
331 The configuration file is saved in filesystem meta-data.
332 The snapshots directory defaults to
333 .Pa /var/hammer/<pfs>
334 .Pa ( /var/hammer/root
338 The format of the configuration file is:
339 .Bd -literal -offset indent
340 snapshots <period> <retention-time> [any]
341 prune <period> <max-runtime>
342 rebalance <period> <max-runtime>
343 reblock <period> <max-runtime>
344 recopy <period> <max-runtime>
348 .Bd -literal -offset indent
349 snapshots 1d 60d # 0d 0d for PFS /tmp, /var/tmp, /usr/obj
356 Time is given with a suffix of
362 meaning day, hour, minute and second.
366 directive has a period of 0 and a retention time of 0
367 then snapshot generation is disabled, removal of old snapshots are
368 disabled, and prunes will use
369 .Cm prune-everything .
372 directive has a period of 0 but a non-zero retention time
373 then this command will not create any new snapshots but will remove old
374 snapshots it finds based on the retention time.
376 By default only snapshots in the form
377 .Ql snap- Ns Ar yyyymmdd Ns Op - Ns Ar HHMM
381 directive is specified as a third argument on the
383 config line then any softlink of the form
384 .Ql *- Ns Ar yyyymmdd Ns Op - Ns Ar HHMM
386 .Ql *. Ns Ar yyyymmdd Ns Op - Ns Ar HHMM
389 A prune max-runtime of 0 means unlimited.
391 If period hasn't passed since the previous
394 For example a day has passed when midnight is passed (localtime).
402 The default configuration file will create a daily snapshot, do a daily
403 pruning, rebalancing and reblocking run and a monthly recopy run.
404 Reblocking is defragmentation with a level of 95%,
405 and recopy is full defragmentation.
407 By default prune and rebalance operations are time limited to 5 minutes,
408 reblock operations to a bit over 5 minutes,
409 and recopy operations to a bit over 10 minutes.
410 Reblocking and recopy runs are each broken down into four separate functions:
411 btree, inodes, dirs and data.
412 Each function is time limited to the time given in the configuration file,
413 but the btree, inodes and dirs functions usually does not take very long time,
414 full defragmentation is always used for these three functions.
415 Also note that this directive will by default disable snapshots on
422 The defaults may be adjusted by modifying the configuration file.
423 The pruning and reblocking commands automatically maintain a cyclefile
424 for incremental operation.
425 If you interrupt (^C) the program the cyclefile will be updated,
427 may continue to run in the background for a few seconds until the
429 ioctl detects the interrupt.
432 PFS option can be set to use another location for the snapshots directory.
434 Work on this command is still in progress.
436 An ability to remove snapshots dynamically as the
437 file system becomes full.
439 .It Cm config Op Ar filesystem Op Ar configfile
442 If zero or one arguments are specified this function dumps the current
443 configuration file to stdout.
444 Zero arguments specifies the PFS containing the current directory.
445 This configuration file is stored in filesystem meta-data.
446 If two arguments are specified this function installs a new config file.
450 versions less than 3 the configuration file is by default stored in
451 .Pa <pfs>/snapshots/config ,
452 but in all later versions the configuration file is stored in filesystem
454 .\" ==== viconfig ====
455 .It Cm viconfig Op Ar filesystem
458 Edit the configuration file and reinstall into filesystem meta-data when done.
459 Zero arguments specifies the PFS containing the current directory.
461 .It Cm expand Ar filesystem Ar device
462 This command will format
464 and add all of its space to
468 All existing data contained on
470 will be destroyed by this operation!
475 filesystem, formatting will be denied.
476 You can overcome this sanity check
479 to erase the beginning sectors of the device.
480 Also remember that you have to specify
482 together with any other device that make the filesystem, colon-separated to
484 .\" ==== snapshot ====
485 .It Cm snapshot Oo Ar filesystem Oc Ar snapshot-dir
486 .It Cm snapshot Ar filesystem Ar snapshot-dir Op Ar note
487 Takes a snapshot of the file system either explicitly given by
489 or implicitly derived from the
491 argument and creates a symlink in the directory provided by
493 pointing to the snapshot.
496 is not a directory, it is assumed to be a format string passed to
498 with the current time as parameter.
501 refers to an existing directory, a default format string of
503 is assumed and used as name for the newly created symlink.
505 Snapshot is a per PFS operation, so a
507 file system and each PFS in it have to be snapshot separately.
509 Example, assuming that
517 are file systems on their own, the following invocations:
518 .Bd -literal -offset indent
519 hammer snapshot /mysnapshots
521 hammer snapshot /mysnapshots/%Y-%m-%d
523 hammer snapshot /obj /mysnapshots/obj-%Y-%m-%d
525 hammer snapshot /usr /my/snaps/usr "note"
528 Would create symlinks similar to:
529 .Bd -literal -offset indent
530 /mysnapshots/snap-20080627-1210 -> /@@0x10d2cd05b7270d16
532 /mysnapshots/2008-06-27 -> /@@0x10d2cd05b7270d16
534 /mysnapshots/obj-2008-06-27 -> /obj@@0x10d2cd05b7270d16
536 /my/snaps/usr/snap-20080627-1210 -> /usr@@0x10d2cd05b7270d16
541 version 3+ filesystem the snapshot is also recorded in meta-data
542 along with the optional
548 .It Cm snap Ar path Op Ar note
551 Create a snapshot for the PFS containing
553 and create a snapshot softlink.
554 If the path specified is a
555 directory a standard snapshot softlink will be created in the directory.
556 The snapshot softlink points to the base of the mounted PFS.
557 .It Cm snaplo Ar path Op Ar note
560 Create a snapshot for the PFS containing
562 and create a snapshot softlink.
563 If the path specified is a
564 directory a standard snapshot softlink will be created in the directory.
565 The snapshot softlink points into the directory it is contained in.
566 .It Cm snapq Ar dir Op Ar note
569 Create a snapshot for the PFS containing the specified directory but do
570 not create a softlink.
571 Instead output a path which can be used to access
572 the directory via the snapshot.
574 An absolute or relative path may be specified.
575 The path will be used as-is as a prefix in the path output to stdout.
577 snap and snapshot directives the snapshot transaction id will be registered
578 in the filesystem meta-data.
579 .It Cm snaprm Bro Ar path | transid Brc Ar ...
582 Remove a snapshot given its softlink or transaction id.
583 If specifying a transaction id
584 the snapshot is removed from filesystem meta-data but you are responsible
585 for removing any related softlinks.
586 .It Cm snapls Op Ar path ...
589 Dump the snapshot meta-data for PFSs containing each
591 listing all available snapshots and their notes.
592 If no arguments are specified snapshots for the PFS containing the
593 current directory are listed.
594 This is the definitive list of snapshots for the filesystem.
596 .It Cm prune Ar softlink-dir
597 Prune the file system based on previously created snapshot softlinks.
598 Pruning is the act of deleting file system history.
601 command will delete file system history such that
602 the file system state is retained for the given snapshots,
603 and all history after the latest snapshot.
604 By setting the per PFS parameter
606 history is guaranteed to be saved at least this time interval.
607 All other history is deleted.
609 The target directory is expected to contain softlinks pointing to
610 snapshots of the file systems you wish to retain.
611 The directory is scanned non-recursively and the mount points and
612 transaction ids stored in the softlinks are extracted and sorted.
613 The file system is then explicitly pruned according to what is found.
614 Cleaning out portions of the file system is as simple as removing a
615 snapshot softlink and then running the
619 As a safety measure pruning only occurs if one or more softlinks are found
622 snapshot id extension.
623 Currently the scanned softlink directory must contain softlinks pointing
627 The softlinks may specify absolute or relative paths.
628 Softlinks must use 20-character
630 transaction ids, as might be returned from
631 .Nm Cm synctid Ar filesystem .
633 Pruning is a per PFS operation, so a
635 file system and each PFS in it have to be pruned separately.
637 Note that pruning a file system may not immediately free-up space,
638 though typically some space will be freed if a large number of records are
640 The file system must be reblocked to completely recover all available space.
642 Example, lets say your that you didn't set
644 and snapshot directory contains the following links:
645 .Bd -literal -offset indent
646 lrwxr-xr-x 1 root wheel 29 May 31 17:57 snap1 ->
647 /usr/obj/@@0x10d2cd05b7270d16
649 lrwxr-xr-x 1 root wheel 29 May 31 17:58 snap2 ->
650 /usr/obj/@@0x10d2cd13f3fde98f
652 lrwxr-xr-x 1 root wheel 29 May 31 17:59 snap3 ->
653 /usr/obj/@@0x10d2cd222adee364
656 If you were to run the
658 command on this directory, then the
661 mount will be pruned to retain the above three snapshots.
662 In addition, history for modifications made to the file system older than
663 the oldest snapshot will be destroyed and history for potentially fine-grained
664 modifications made to the file system more recently than the most recent
665 snapshot will be retained.
667 If you then delete the
669 softlink and rerun the
672 history for modifications pertaining to that snapshot would be destroyed.
676 filesystem versions 3+ this command also scans the snapshots stored
677 in the filesystem meta-data and includes them in the prune.
678 .\" ==== prune-everything ====
679 .It Cm prune-everything Ar filesystem
680 This command will remove all historical records from the file system.
681 This directive is not normally used on a production system.
683 This command does not remove snapshot softlinks but will delete all
684 snapshots recorded in filesystem meta-data (for filesystem version 3+).
685 The user is responsible for deleting any softlinks.
686 .\" ==== rebalance ====
687 .It Cm rebalance Ar filesystem Op Ar saturation_level
688 This command will rebalance the B-tree, nodes with small number of
689 elements will be combined and element counts will be smoothed out
692 The saturation level is a percentage between 50 and 100.
693 The default is 75 percent.
694 .\" ==== reblock* ====
695 .It Cm reblock Ar filesystem Op Ar fill_percentage
696 .It Cm reblock-btree Ar filesystem Op Ar fill_percentage
697 .It Cm reblock-inodes Ar filesystem Op Ar fill_percentage
698 .It Cm reblock-dirs Ar filesystem Op Ar fill_percentage
699 .It Cm reblock-data Ar filesystem Op Ar fill_percentage
700 Attempt to defragment and free space for reuse by reblocking a live
703 Big-blocks cannot be reused by
705 until they are completely free.
706 This command also has the effect of reordering all elements, effectively
707 defragmenting the file system.
709 The default fill percentage is 100% and will cause the file system to be
710 completely defragmented.
711 All specified element types will be reallocated and rewritten.
712 If you wish to quickly free up space instead try specifying
713 a smaller fill percentage, such as 90% or 80% (the
715 suffix is not needed).
717 Since this command may rewrite the entire contents of the disk it is
718 best to do it incrementally from a
724 options to limit the run time.
725 The file system would thus be defragmented over long period of time.
727 It is recommended that separate invocations be used for each data type.
728 B-tree nodes, inodes, and directories are typically the most important
729 elements needing defragmentation.
730 Data can be defragmented over a longer period of time.
732 Reblocking is a per PFS operation, so a
734 file system and each PFS in it have to be reblocked separately.
735 .\" ==== pfs-status ====
736 .It Cm pfs-status Ar dirpath ...
737 Retrieve the mirroring configuration parameters for the specified
739 file systems or pseudo-filesystems (PFS's).
740 .\" ==== pfs-master ====
741 .It Cm pfs-master Ar dirpath Op Ar options
742 Create a pseudo-filesystem (PFS) inside a
745 Up to 65535 such file systems can be created.
746 Each PFS uses an independent inode numbering space making it suitable
747 for use as a replication source or target.
751 directive creates a PFS that you can read, write, and use as a mirroring
754 It is recommended to use a
756 mount to access a PFS, for more information see
758 .\" ==== pfs-slave ====
759 .It Cm pfs-slave Ar dirpath Op Ar options
760 Create a pseudo-filesystem (PFS) inside a
763 Up to 65535 such file systems can be created.
764 Each PFS uses an independent inode numbering space making it suitable
765 for use as a replication source or target.
769 directive creates a PFS that you can use as a mirroring target.
770 You will not be able to access a slave PFS until you have completed the
771 first mirroring operation with it as the target (its root directory will
772 not exist until then).
774 Access to the pfs-slave via the special softlink, as described in the
779 dynamically modify the snapshot transaction id by returning a dynamic result
784 A PFS can only be truly destroyed with the
787 Removing the softlink will not destroy the underlying PFS.
789 It is recommended to use a
791 mount to access a PFS, for more information see
793 .\" ==== pfs-update ====
794 .It Cm pfs-update Ar dirpath Op Ar options
795 Update the configuration parameters for an existing
797 file system or pseudo-filesystem.
798 Options that may be specified:
799 .Bl -tag -width indent
800 .It Cm sync-beg-tid= Ns Ar 0x16llx
801 This is the automatic snapshot access starting transaction id for
803 This parameter is normally updated automatically by the
807 It is important to note that accessing a mirroring slave
808 with a transaction id greater than the last fully synchronized transaction
809 id can result in an unreliable snapshot since you will be accessing
810 data that is still undergoing synchronization.
812 Manually modifying this field is dangerous and can result in a broken mirror.
813 .It Cm sync-end-tid= Ns Ar 0x16llx
814 This is the current synchronization point for mirroring slaves.
815 This parameter is normally updated automatically by the
819 Manually modifying this field is dangerous and can result in a broken mirror.
820 .It Cm shared-uuid= Ns Ar uuid
821 Set the shared UUID for this file system.
822 All mirrors must have the same shared UUID.
823 For safety purposes the
825 directives will refuse to operate on a target with a different shared UUID.
827 Changing the shared UUID on an existing, non-empty mirroring target,
828 including an empty but not completely pruned target,
829 can lead to corruption of the mirroring target.
830 .It Cm unique-uuid= Ns Ar uuid
831 Set the unique UUID for this file system.
832 This UUID should not be used anywhere else,
833 even on exact copies of the file system.
834 .It Cm label= Ns Ar string
835 Set a descriptive label for this file system.
836 .It Cm snapshots= Ns Ar string
837 Specify the snapshots directory which
840 will use to manage this PFS.
841 .Bl -tag -width indent
842 .It Nm HAMMER No version 2-
843 The snapshots directory does not need to be configured for
844 PFS masters and will default to
845 .Pa <pfs>/snapshots .
847 PFS slaves are mirroring slaves so you cannot configure a snapshots
848 directory on the slave itself to be managed by the slave's machine.
849 In fact, the slave will likely have a
851 sub-directory mirrored
852 from the master, but that directory contains the configuration the master
853 is using for its copy of the file system, not the configuration that we
854 want to use for our slave.
856 It is recommended that
857 .Pa <fs>/var/slaves/<name>
858 be configured for a PFS slave, where
864 is an appropriate label.
865 .It Nm HAMMER No version 3+
866 The snapshots directory does not need to be configured for PFS masters or
868 The snapshots directory defaults to
869 .Pa /var/hammer/<pfs>
870 .Pa ( /var/hammer/root
874 You can control snapshot retention on your slave independent of the master.
875 .It Cm snapshots-clear
878 directory path for this PFS.
879 .It Cm prune-min= Ns Ar N Ns Cm d
880 .It Cm prune-min= Ns Oo Ar N Ns Cm d/ Oc Ns \
881 Ar hh Ns Op Cm \&: Ns Ar mm Ns Op Cm \&: Ns Ar ss
882 Set the minimum fine-grained data retention period.
884 always retains fine-grained history up to the most recent snapshot.
885 You can extend the retention period further by specifying a non-zero
887 Any snapshot softlinks within the retention period are ignored
888 for the purposes of pruning (the fine grained history is retained).
889 Number of days, hours, minutes and seconds are given as
894 Because the transaction id in the snapshot softlink cannot be used
895 to calculate a timestamp,
897 uses the earlier of the
901 field of the softlink to
902 determine which snapshots fall within the retention period.
903 Users must be sure to retain one of these two fields when manipulating
906 .\" ==== pfs-upgrade ====
907 .It Cm pfs-upgrade Ar dirpath
908 Upgrade a PFS from slave to master operation.
909 The PFS will be rolled back to the current end synchronization transaction id
910 (removing any partial synchronizations), and will then become writable.
914 currently supports only single masters and using
915 this command can easily result in file system corruption
916 if you don't know what you are doing.
918 This directive will refuse to run if any programs have open descriptors
919 in the PFS, including programs chdir'd into the PFS.
920 .\" ==== pfs-downgrade ====
921 .It Cm pfs-downgrade Ar dirpath
922 Downgrade a master PFS from master to slave operation
923 The PFS becomes read-only and access will be locked to its
926 This directive will refuse to run if any programs have open descriptors
927 in the PFS, including programs chdir'd into the PFS.
928 .\" ==== pfs-destroy ====
929 .It Cm pfs-destroy Ar dirpath
930 This permanently destroys a PFS.
932 This directive will refuse to run if any programs have open descriptors
933 in the PFS, including programs chdir'd into the PFS.
934 .\" ==== mirror-read ====
935 .It Cm mirror-read Ar filesystem Op Ar begin-tid
936 Generate a mirroring stream to stdout.
937 The stream ends when the transaction id space has been exhausted.
938 .\" ==== mirror-read-stream ====
939 .It Cm mirror-read-stream Ar filesystem Op Ar begin-tid
940 Generate a mirroring stream to stdout.
941 Upon completion the stream is paused until new data is synced to the
942 master, then resumed.
943 Operation continues until the pipe is broken.
944 .\" ==== mirror-write ====
945 .It Cm mirror-write Ar filesystem
946 Take a mirroring stream on stdin.
948 This command will fail if the
950 configuration field for the two file systems do not match.
952 If the target PFS does not exist this command will ask you whether
953 you want to create a compatible PFS slave for the target or not.
954 .\" ==== mirror-dump ====
960 to dump an ASCII representation of the mirroring stream.
961 .\" ==== mirror-copy ====
962 .\".It Cm mirror-copy Ar [[user@]host:]filesystem [[user@]host:]filesystem
964 Oo Oo Ar user Oc Ns Cm @ Ns Ar host Ns Cm \&: Oc Ns Ar filesystem \
965 Oo Oo Ar user Oc Ns Cm @ Ns Ar host Ns Cm \&: Oc Ns Ar filesystem
966 This is a shortcut which pipes a
971 If a remote host specification is made the program forks a
977 on the appropriate host.
978 The source may be a master or slave PFS, and the target must be a slave PFS.
980 This command also established full duplex communication and turns on
981 the two-way protocol feature which automatically negotiates transaction id
982 ranges without having to use a cyclefile.
983 If the operation completes successfully the target PFS's
986 Note that you must re-chdir into the target PFS to see the updated information.
987 If you do not you will still be in the previous snapshot.
989 If the target PFS does not exist this command will ask you whether
990 you want to create a compatible PFS slave for the target or not.
991 .\" ==== mirror-stream ====
992 .\".It Cm mirror-stream Ar [[user@]host:]filesystem [[user@]host:]filesystem
993 .It Cm mirror-stream \
994 Oo Oo Ar user Oc Ns Cm @ Ns Ar host Ns Cm \&: Oc Ns Ar filesystem \
995 Oo Oo Ar user Oc Ns Cm @ Ns Ar host Ns Cm \&: Oc Ns Ar filesystem
996 This command works similarly to
998 but does not exit after the initial mirroring completes.
999 The mirroring operation will resume as changes continue to be made to the
1001 The command is commonly used with
1005 options to keep the mirroring target in sync with the source on a continuing
1008 If the pipe is broken the command will automatically retry after sleeping
1010 The time slept will be 15 seconds plus the time given in the
1014 This command also detects the initial-mirroring case and spends some
1015 time scanning the B-Tree to find good break points, allowing the initial
1016 bulk mirroring operation to be broken down into about 20 separate pieces.
1017 This means that the user can kill and restart the operation and it will
1018 not have to start from scratch once it has gotten past the first chunk.
1021 option may be used to disable this feature and perform an initial bulk
1023 .\" ==== version ====
1024 .It Cm version Ar filesystem
1025 This command returns the
1027 file system version for the specified
1029 as well as the range of versions supported in the kernel.
1032 option may be used to remove the summary at the end.
1033 .\" ==== version-upgrade ====
1034 .It Cm version-upgrade Ar filesystem Ar version Op Cm force
1035 This command upgrades the
1040 Once upgraded a file system may not be downgraded.
1041 If you wish to upgrade a file system to a version greater or equal to the
1042 work-in-progress version number you must specify the
1045 Use of WIP versions should be relegated to testing and may require wiping
1046 the filesystem as development progresses, even though the WIP version might
1050 This command operates on the entire
1052 file system and is not a per PFS operation.
1053 All PFS's will be affected.
1054 .Bl -tag -width indent
1057 default version, first
1062 default version, new directory entry layout.
1063 This version is using a new directory hash key.
1066 New snapshot management, using filesystem meta-data for saving
1067 configuration file and snapshots (transaction ids etc.).
1068 Also default snapshots directory has changed.
1071 New REDO, faster flush/sync.
1076 .Sh PSEUDO-FILESYSTEM (PFS) NOTES
1077 The root of a PFS is not hooked into the primary
1079 file system as a directory.
1082 creates a special softlink called
1084 (exactly 10 characters long) in the primary
1088 then modifies the contents of the softlink as read by
1090 and thus what you see with an
1092 command or if you were to
1095 If the PFS is a master the link reflects the current state of the PFS.
1096 If the PFS is a slave the link reflects the last completed snapshot, and the
1097 contents of the link will change when the next snapshot is completed, and
1102 utility employs numerous safeties to reduce user foot-shooting.
1105 directive requires that the target be configured as a slave and that the
1107 field of the mirroring source and target match.
1108 .Sh UPGRADE INSTRUCTIONS HAMMER V1 TO V2
1109 This upgrade changes the way directory entries are stored.
1110 It is possible to upgrade a V1 filesystem to V2 in place, but
1111 directories created prior to the upgrade will continue to use
1114 Note that the slave mirroring code in the target kernel had bugs in
1115 V1 which can create an incompatible root directory on the slave.
1118 master created after the upgrade with a
1120 slave created prior to the upgrade.
1122 Any directories created after upgrading will use a new layout.
1123 .Sh UPGRADE INSTRUCTIONS HAMMER V2 TO V3
1124 This upgrade adds meta-data elements to the B-Tree.
1125 It is possible to upgrade a V2 filesystem to V3 in place.
1126 After issuing the upgrade be sure to run a
1129 to perform post-upgrade tasks.
1131 After making this upgrade running a
1136 directory for each PFS mount into
1137 .Pa /var/hammer/<pfs> .
1140 root mount will migrate
1143 .Pa /var/hammer/root .
1144 Migration occurs only once and only if you have not specified
1145 a snapshots directory in the PFS configuration.
1146 If you have specified a snapshots directory in the PFS configuration no
1147 automatic migration will occur.
1149 For slaves, if you desire, you can migrate your snapshots
1150 config to the new location manually and then clear the
1151 snapshot directory configuration in the slave PFS.
1152 The new snapshots hierarchy is designed to work with
1153 both master and slave PFSs equally well.
1155 In addition, the old config file will be moved to meta-data,
1156 editable via the new
1160 The old config file will be deleted.
1161 Migration occurs only once.
1163 The V3 filesystem has new
1165 directives for creating snapshots.
1166 All snapshot directives, including the original, will create
1167 meta-data entries for the snapshots and the pruning code will
1168 automatically incorporate these entries into its list and
1169 expire them the same way it expires softlinks.
1170 If you by accident blow away your snapshot softlinks you can use the
1172 directive to get a definitive list from the meta-data and
1173 regenerate them from that list.
1178 to backup filesystems your scripts may be using the
1180 directive to generate transaction ids.
1181 This directive does not create a snapshot.
1182 You will have to modify your scripts to use the
1184 directive to generate the linkbuf for the softlink you create, or
1185 use one of the other
1190 directive will continue to work as expected and in V3 it will also
1191 record the snapshot transaction id in meta-data.
1192 You may also want to make use of the new
1194 tag for the meta-data.
1197 If you used to remove snapshot softlinks with
1199 you should probably start using the
1201 directive instead to also remove the related meta-data.
1202 The pruning code scans the meta-data so just removing the
1203 softlink is not sufficient.
1207 .Bl -tag -width ".It Pa <fs>/var/slaves/<name>" -compact
1208 .It Pa <pfs>/snapshots
1209 default per PFS snapshots directory
1212 .It Pa /var/hammer/<pfs>
1213 default per PFS snapshots directory (not root)
1216 .It Pa /var/hammer/root
1217 default snapshots directory for root directory
1220 .It Pa <snapshots>/config
1227 .It Pa <fs>/var/slaves/<name>
1228 recommended slave PFS snapshots directory
1236 .Xr periodic.conf 5 ,
1237 .Xr mount_hammer 8 ,
1243 utility first appeared in
1246 .An Matthew Dillon Aq dillon@backplane.com