xref: /dragonfly/sbin/jscan/jscan.8 (revision b7367ef6)
1.\" Copyright (c) 2004,2005 The DragonFly Project.  All rights reserved.
2.\"
3.\" This code is derived from software contributed to The DragonFly Project
4.\" by Matthew Dillon <dillon@backplane.com>
5.\"
6.\"
7.\" Redistribution and use in source and binary forms, with or without
8.\" modification, are permitted provided that the following conditions
9.\" are met:
10.\"
11.\" 1. Redistributions of source code must retain the above copyright
12.\"    notice, this list of conditions and the following disclaimer.
13.\" 2. Redistributions in binary form must reproduce the above copyright
14.\"    notice, this list of conditions and the following disclaimer in
15.\"    the documentation and/or other materials provided with the
16.\"    distribution.
17.\" 3. Neither the name of The DragonFly Project nor the names of its
18.\"    contributors may be used to endorse or promote products derived
19.\"    from this software without specific, prior written permission.
20.\"
21.\" THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
22.\" ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
23.\" LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
24.\" FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE
25.\" COPYRIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
26.\" INCIDENTAL, SPECIAL, EXEMPLARY OR CONSEQUENTIAL DAMAGES (INCLUDING,
27.\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
28.\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
29.\" AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
30.\" OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
31.\" OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32.\" SUCH DAMAGE.
33.\"
34.\" $DragonFly: src/sbin/jscan/jscan.8,v 1.17 2007/09/14 23:47:53 swildner Exp $
35.\"
36.Dd March 6, 2005
37.Dt JSCAN 8
38.Os
39.Sh NAME
40.Nm jscan
41.Nd journal file processing program
42.Sh SYNOPSIS
43.Nm
44.Fl 2dfuvF
45.Op Fl c Ar count[k,m,g,t]
46.Op Fl D Ar directory
47.Op Fl m Ar mirror_transid_file/none
48.Op Fl o/O Ar output_transid_file/none
49.Op Fl s Ar size[k,m,g,t]
50.Op Fl w/W Ar journal_prefix
51.Op Ar journal_prefix/file
52.Sh DESCRIPTION
53The
54.Nm
55utility scans journal file or input stream for the purposes of debugging
56dumps, restoration, undo, mirroring, and other journaling features.
57.Bl -tag -width indent
58.It Fl 2
59Implement the full-duplex acknowledgement protocol on the input descriptor.
60Note that shell pipes are full-duplex and can be used with this option.
61.It Fl c Ar count
62Specify the number of transaction records which should be scanned, then exit.
63This option is typically used along with
64.Fl m
65to limit the amount of work that
66.Nm
67does, giving you the ability to incrementally run a mirror forwards or
68backwards.  It is not usually used when piping in a live journal, but it
69can be.
70.It Fl d
71Display the contents of the journaling file or stream in a human readable
72format on stderr.  Note that stdout is used only for
73.Fl o .
74.It Fl f
75.Nm
76will sleep for 5 seconds and loop when it hits EOF on file or prefix
77set input rather than exit.  This option is typically used when running
78on an input file or prefix set which is live (being written to by
79another
80.Nm
81instance).
82.It Fl D Ar directory
83Specify the base directory for the mirroring option.
84.It Fl m Ar mirror_transid_file/none
85Generate a mirror in the directory specified by
86.Fl D
87or, if not specified, the current directory.
88The
89.Ar mirror_transid_file
90will be used to track the transaction id representing the current
91synchronization point for the mirror.  The keyword
92.Ar none
93may be specified if no tracking file is desired.  However, if no tracking
94file is specified it will not be possible to roll the mirror forwards or
95backwards or restart the journaling stream being used to generate the mirror.
96.Pp
97It is important to note that journaling streams can contain meta-transactions
98representing huge, multi-gigabyte operations.  If the journaling data is
99not being recorded to regular files via
100.Fl w/W
101it is possible that
102.Nm
103could run itself out of memory trying to record the meta-transactions.
104In addition, the mirror would not be restartable.  If the journaling data
105is being recorded via
106.Fl w/W
107and a mirroring transaction id file is being kept, the mirror can be
108restarted.
109.Pp
110While it is possible to run a journaling stream directly into a mirror,
111it is more typical to file the jornaling stream with
112.Fl w
113and catch the mirror up as a batch job with the journaling file set prefix
114specified as the input every so often.  This way the system operator can
115use other
116.Nm
117commands to, for example, run a mirror backwards and forwards in time.
118.It Fl o/O Ar output_transid_file/none
119Generate a journaling stream on stdout using the specified file to track
120the transaction id to help with restarts.
121The
122.Fl o
123option indicates a half-duplex output stream while the
124.Fl O
125option indicates a full-duplex (ACK protocol) output stream.
126.Pp
127This option is not really designed to output to regular files because it
128does NOT necessarily weed out duplicate records.  When both the input
129stream and output stream are full-duplex and
130.Fl w/W
131is not specified,
132.Nm
133acts as a stateless transceiver and the input stream is not acked until
134an ack is received from the output stream.
135.Pp
136This option is most typically used in conjunction with
137.Fl w/W .
138In this case the ACK protocol is handled independently for the input side
139and the output side uses the journaling data recorded by
140.Fl w/W
141as a buffer.
142.Pp
143In half-duplex output mode the output transaction id file is updated
144after a raw transaction record has been successfully written to stdout.
145In full-duplex output mode the file is only updated with ACK data returned
146on the stdout descriptor.
147.Pp
148As with the
149.Fl m
150option, you can combine
151.Fl o
152in a journaling pipe with other options, but if you are trying to use it
153as a buffer it may be better to have it separately pull its data off of
154a journaling file set generated via
155.Fl w .
156.It Fl s Ar size
157Change the size limit for rotating files created via
158.Fl w .
159The default is 100M.  Values are in bytes or may be suffixed with k,
160m, or g.
161If a raw transaction causes the file's size limit to be exceeded, a new file
162will be created.  If a raw transaction is, in-whole, larger the the file's
163size limit, the raw transaction will still be fully written to the file before
164a new file is created.  Raw transactions are typically limited to the size
165of the source system's memory FIFO.  This option is typically used to size
166journaling files to fit onto the appropriate backup media or to provide
167bite-sized chunks for other programs to ingest.
168.Pp
169When restarting a journal, a new sequence number will always be chosen for
170the resumption of data recording.  No existing file will be appended to when
171.Nm
172is reinvoked.
173.It Fl u
174Will cause the journal to be scanned backwards (requires seekable media).
175Transactions will be dumped in reverse order.  If mirroring, the UNDO
176data will be executed.  If not specified, 1 hour's worth of data will be
177undone.  Can only be used with a journaling file or journaling prefix
178as the input.
179.It Fl v
180Increase verbosity on stderr.  This option is primarily used for debugging.
181.It Fl w Ar prefix
182The received journaling stream is recorded in journaling files named
183.Ar <prefix>.<seq>
184and the current transaction id is tracked in a file named
185.Ar <prefix>.transid .
186A journaling file is closed out and a new file with the next sequence
187number is created once the file surpasses 100MB.
188.Pp
189This option is robust across restarts.  The current transaction id
190will be read and the input stream will be skipped until it is reached.
191If the input is a journaling file or prefix set,
192.Nm
193will be able to quickly seek to the restart point.
194.Pp
195NOTE: If
196you are generating a mirror with the same command via
197.Fl m ,
198and the journaling data input is a stream rather than a file or prefix
199set, you must use
200.Fl w/W
201if you want the mirror to be restartable.  This is because while we can
202pick up the transaction id where we left off, that raw transaction id may
203have cut a larger meta-transaction in half and the mirroring code will
204not be able to access the whole of the transaction unless it has a file
205or prefix set to work with.
206.It Fl W Ar prefix
207Similar to
208.Fl w
209except that the journaling files created are strictly temporary and will
210be deleted once they exceed the size limit AND the related meta-transactions
211have been completed.
212.Pp
213If combined with
214.Fl m ,
215the meta-transactions are considered to be completed only when the mirror
216finishes executing them.  It is possible for several sequence number files
217to build if a particularly large meta-transaction is coming down the pipe.
218.Pp
219If combined with
220.Fl o/O ,
221the meta-transactions are considered to be completed when the data has
222been successfully written out to the pipe in half duplex mode, or when
223the ACK has been received in full-duplex mode.
224.Pp
225If both
226.Fl m
227and
228.Fl w/W
229is used, the journaling data files are only deleted when both actions
230no longer need the data.
231.It Fl F
232Forces
233.Nm
234to fsync() after updating a journaling file prior to acknowledging the
235data or updating a transaction-id-tracking file.  If specified twice,
236.Nm
237will also fsync() after updating the transaction-id-tracking file.
238.It Ar journal_prefix/file
239Specify the input to
240.Nm .
241This can be a journaling file set prefix
242or it can be a plain file.  If no input file is specified, stdin is
243assumed.  Note that when generating a mirror from a stdin stream, the
244mirror will not be restartable unless
245.Fl w/W
246is also used.
247.El
248.Sh OPERATIONAL NOTES
249It is often important to be able to quickly stage journaled data through
250a dedicated backup machine on a LAN.  There are several places where data
251can be buffered and staged out.
252.Pp
253The machine generating the journal typically buffers several megabytes of
254journal data in the kernel.  This local machine can pipe that data to
255.Nm
256or some other locally run program to add another buffering stage, or you
257can directly attach a TCP connection to the kernel's journaling output.
258.Pp
259The LAN backup box typically buffers gigabytes worth of data by running
260multiple jscans.  The
261.Nm
262on the receiving end of the TCP or pipe (for
263example, via ssh) typically records the data via the
264.Fl w
265option, and then runs other
266.Nm
267programs from scripts or cron to take that data and copy it to your
268off-site backup machine.  Other
269.Nm
270programs may use the same data
271set to generate mirrors or other backup streams.
272.Pp
273It should be noted that if
274.Fl w/W
275is specified, both mirroring mode and output mode will internally
276fork the program once the appropriate synchronization point has been reached,
277effectively decoupling their operation, and read all of their data via
278the journaling files written out by the master program.  In particular,
279blockages in the mirroring and output code will not affect our ability
280to buffer the journaling input data via
281.Fl w/W .
282If
283.Fl w/W
284is not specified then neither the mirroring or output modes will fork.  Under
285these conditions, if the input is a stream rather than a file
286.Nm
287will be forced to buffer meta-transactions (for mirroring) entirely in
288memory, which could present a serious problem since a single meta-transaction
289can exceed a gigabyte (e.g. if someone were to do a single write() system
290call writing a gigabyte all in one go).
291.Sh SEE ALSO
292.Xr mountctl 8
293.Sh CAVEATS
294This utility is currently under construction and not all features have been
295implemented yet.
296In fact, most have not.
297.Sh HISTORY
298The
299.Nm
300utility first appeared in
301.Dx 1.3 .
302