1GIT-PACK-OBJECTS(1) Git Manual GIT-PACK-OBJECTS(1)
2
6 git-pack-objects - Create a packed archive of objects
7
9 git pack-objects [-q | --progress | --all-progress] [--all-progress-implied]
10 [--no-reuse-delta] [--delta-base-offset] [--non-empty]
11 [--local] [--incremental] [--window=<n>] [--depth=<n>]
12 [--revs [--unpacked | --all]] [--keep-pack=<pack-name>]
13 [--cruft] [--cruft-expiration=<time>]
14 [--stdout [--filter=<filter-spec>] | <base-name>]
15 [--shallow] [--keep-true-parents] [--[no-]sparse] < <object-list>
16
18 Reads list of objects from the standard input, and writes either one or
19 more packed archives with the specified base-name to disk, or a packed
20 archive to the standard output.
21 A packed archive is an efficient way to transfer a set of objects
23 between two repositories as well as an access efficient archival
24 format. In a packed archive, an object is either stored as a compressed
25 whole or as a difference from some other object. The latter is often
26 called a delta.
27 The packed archive format (.pack) is designed to be self-contained so
29 that it can be unpacked without any further information. Therefore,
30 each object that a delta depends upon must be present within the pack.
31 A pack index file (.idx) is generated for fast, random access to the
33 objects in the pack. Placing both the index file (.idx) and the packed
34 archive (.pack) in the pack/ subdirectory of $GIT_OBJECT_DIRECTORY (or
35 any of the directories on $GIT_ALTERNATE_OBJECT_DIRECTORIES) enables
36 Git to read from the pack archive.
37 The git unpack-objects command can read the packed archive and expand
39 the objects contained in the pack into "one-file one-object" format;
40 this is typically done by the smart-pull commands when a pack is
41 created on-the-fly for efficient network transport by their peers.
42
44 base-name
45 Write into pairs of files (.pack and .idx), using <base-name> to
46 determine the name of the created file. When this option is used,
47 the two files in a pair are written in
48 <base-name>-<SHA-1>.{pack,idx} files. <SHA-1> is a hash based on
49 the pack content and is written to the standard output of the
50 command.
51 --stdout
53 Write the pack contents (what would have been written to .pack
54 file) out to the standard output.
55 --revs
57 Read the revision arguments from the standard input, instead of
58 individual object names. The revision arguments are processed the
59 same way as git rev-list with the --objects flag uses its commit
60 arguments to build the list of objects it outputs. The objects on
61 the resulting list are packed. Besides revisions, --not or
62 --shallow <SHA-1> lines are also accepted.
63 --unpacked
65 This implies --revs. When processing the list of revision arguments
66 read from the standard input, limit the objects packed to those
67 that are not already packed.
68 --all
70 This implies --revs. In addition to the list of revision arguments
71 read from the standard input, pretend as if all refs under refs/
72 are specified to be included.
73 --include-tag
75 Include unasked-for annotated tags if the object they reference was
76 included in the resulting packfile. This can be useful to send new
77 tags to native Git clients.
78 --stdin-packs
80 Read the basenames of packfiles (e.g., pack-1234abcd.pack) from the
81 standard input, instead of object names or revision arguments. The
82 resulting pack contains all objects listed in the included packs
83 (those not beginning with ^), excluding any objects listed in the
84 excluded packs (beginning with ^).
85 Incompatible with --revs, or options that imply --revs (such as
87 --all), with the exception of --unpacked, which is compatible.
88 --cruft
90 Packs unreachable objects into a separate "cruft" pack, denoted by
91 the existence of a .mtimes file. Typically used by git repack
92 --cruft. Callers provide a list of pack names and indicate which
93 packs will remain in the repository, along with which packs will be
94 deleted (indicated by the - prefix). The contents of the cruft pack
95 are all objects not contained in the surviving packs which have not
96 exceeded the grace period (see --cruft-expiration below), or which
97 have exceeded the grace period, but are reachable from an other
98 object which hasn’t.
99 When the input lists a pack containing all reachable objects (and
101 lists all other packs as pending deletion), the corresponding cruft
102 pack will contain all unreachable objects (with mtime newer than
103 the --cruft-expiration) along with any unreachable objects whose
104 mtime is older than the --cruft-expiration, but are reachable from
105 an unreachable object whose mtime is newer than the
106 --cruft-expiration).
107 Incompatible with --unpack-unreachable, --keep-unreachable,
109 --pack-loose-unreachable, --stdin-packs, as well as any other
110 options which imply --revs.
111 --cruft-expiration=<approxidate>
113 If specified, objects are eliminated from the cruft pack if they
114 have an mtime older than <approxidate>. If unspecified (and given
115 --cruft), then no objects are eliminated.
116 --window=<n>, --depth=<n>
118 These two options affect how the objects contained in the pack are
119 stored using delta compression. The objects are first internally
120 sorted by type, size and optionally names and compared against the
121 other objects within --window to see if using delta compression
122 saves space. --depth limits the maximum delta depth; making it too
123 deep affects the performance on the unpacker side, because delta
124 data needs to be applied that many times to get to the necessary
125 object.
126 The default value for --window is 10 and --depth is 50. The maximum
128 depth is 4095.
129 --window-memory=<n>
131 This option provides an additional limit on top of --window; the
132 window size will dynamically scale down so as to not take up more
133 than <n> bytes in memory. This is useful in repositories with a mix
134 of large and small objects to not run out of memory with a large
135 window, but still be able to take advantage of the large window for
136 the smaller objects. The size can be suffixed with "k", "m", or
137 "g". --window-memory=0 makes memory usage unlimited. The default
138 is taken from the pack.windowMemory configuration variable.
139 --max-pack-size=<n>
141 In unusual scenarios, you may not be able to create files larger
142 than a certain size on your filesystem, and this option can be used
143 to tell the command to split the output packfile into multiple
144 independent packfiles, each not larger than the given size. The
145 size can be suffixed with "k", "m", or "g". The minimum size
146 allowed is limited to 1 MiB. The default is unlimited, unless the
147 config variable pack.packSizeLimit is set. Note that this option
148 may result in a larger and slower repository; see the discussion in
149 pack.packSizeLimit.
150 --honor-pack-keep
152 This flag causes an object already in a local pack that has a .keep
153 file to be ignored, even if it would have otherwise been packed.
154 --keep-pack=<pack-name>
156 This flag causes an object already in the given pack to be ignored,
157 even if it would have otherwise been packed. <pack-name> is the
158 pack file name without leading directory (e.g. pack-123.pack). The
159 option could be specified multiple times to keep multiple packs.
160 --incremental
162 This flag causes an object already in a pack to be ignored even if
163 it would have otherwise been packed.
164 --local
166 This flag causes an object that is borrowed from an alternate
167 object store to be ignored even if it would have otherwise been
168 packed.
169 --non-empty
171 Only create a packed archive if it would contain at least one
172 object.
173 --progress
175 Progress status is reported on the standard error stream by default
176 when it is attached to a terminal, unless -q is specified. This
177 flag forces progress status even if the standard error stream is
178 not directed to a terminal.
179 --all-progress
181 When --stdout is specified then progress report is displayed during
182 the object count and compression phases but inhibited during the
183 write-out phase. The reason is that in some cases the output stream
184 is directly linked to another command which may wish to display
185 progress status of its own as it processes incoming pack data. This
186 flag is like --progress except that it forces progress report for
187 the write-out phase as well even if --stdout is used.
188 --all-progress-implied
190 This is used to imply --all-progress whenever progress display is
191 activated. Unlike --all-progress this flag doesn’t actually force
192 any progress display by itself.
193 -q
195 This flag makes the command not to report its progress on the
196 standard error stream.
197 --no-reuse-delta
199 When creating a packed archive in a repository that has existing
200 packs, the command reuses existing deltas. This sometimes results
201 in a slightly suboptimal pack. This flag tells the command not to
202 reuse existing deltas but compute them from scratch.
203 --no-reuse-object
205 This flag tells the command not to reuse existing object data at
206 all, including non deltified object, forcing recompression of
207 everything. This implies --no-reuse-delta. Useful only in the
208 obscure case where wholesale enforcement of a different compression
209 level on the packed data is desired.
210 --compression=<n>
212 Specifies compression level for newly-compressed data in the
213 generated pack. If not specified, pack compression level is
214 determined first by pack.compression, then by core.compression, and
215 defaults to -1, the zlib default, if neither is set. Add
216 --no-reuse-object if you want to force a uniform compression level
217 on all data no matter the source.
218 --[no-]sparse
220 Toggle the "sparse" algorithm to determine which objects to include
221 in the pack, when combined with the "--revs" option. This algorithm
222 only walks trees that appear in paths that introduce new objects.
223 This can have significant performance benefits when computing a
224 pack to send a small change. However, it is possible that extra
225 objects are added to the pack-file if the included commits contain
226 certain types of direct renames. If this option is not included, it
227 defaults to the value of pack.useSparse, which is true unless
228 otherwise specified.
229 --thin
231 Create a "thin" pack by omitting the common objects between a
232 sender and a receiver in order to reduce network transfer. This
233 option only makes sense in conjunction with --stdout.
234 Note: A thin pack violates the packed archive format by omitting
236 required objects and is thus unusable by Git without making it
237 self-contained. Use git index-pack --fix-thin (see git-index-
238 pack(1)) to restore the self-contained property.
239 --shallow
241 Optimize a pack that will be provided to a client with a shallow
242 repository. This option, combined with --thin, can result in a
243 smaller pack at the cost of speed.
244 --delta-base-offset
246 A packed archive can express the base object of a delta as either a
247 20-byte object name or as an offset in the stream, but ancient
248 versions of Git don’t understand the latter. By default, git
249 pack-objects only uses the former format for better compatibility.
250 This option allows the command to use the latter format for
251 compactness. Depending on the average delta chain length, this
252 option typically shrinks the resulting packfile by 3-5 per-cent.
253 Note: Porcelain commands such as git gc (see git-gc(1)), git repack
255 (see git-repack(1)) pass this option by default in modern Git when
256 they put objects in your repository into pack files. So does git
257 bundle (see git-bundle(1)) when it creates a bundle.
258 --threads=<n>
260 Specifies the number of threads to spawn when searching for best
261 delta matches. This requires that pack-objects be compiled with
262 pthreads otherwise this option is ignored with a warning. This is
263 meant to reduce packing time on multiprocessor machines. The
264 required amount of memory for the delta search window is however
265 multiplied by the number of threads. Specifying 0 will cause Git to
266 auto-detect the number of CPU’s and set the number of threads
267 accordingly.
268 --index-version=<version>[,<offset>]
270 This is intended to be used by the test suite only. It allows to
271 force the version for the generated pack index, and to force 64-bit
272 index entries on objects located above the given offset.
273 --keep-true-parents
275 With this option, parents that are hidden by grafts are packed
276 nevertheless.
277 --filter=<filter-spec>
279 Omits certain objects (usually blobs) from the resulting packfile.
280 See git-rev-list(1) for valid <filter-spec> forms.
281 --no-filter
283 Turns off any previous --filter= argument.
284 --missing=<missing-action>
286 A debug option to help with future "partial clone" development.
287 This option specifies how missing objects are handled.
288 The form --missing=error requests that pack-objects stop with an
290 error if a missing object is encountered. If the repository is a
291 partial clone, an attempt to fetch missing objects will be made
292 before declaring them missing. This is the default action.
293 The form --missing=allow-any will allow object traversal to
295 continue if a missing object is encountered. No fetch of a missing
296 object will occur. Missing objects will silently be omitted from
297 the results.
298 The form --missing=allow-promisor is like allow-any, but will only
300 allow object traversal to continue for EXPECTED promisor missing
301 objects. No fetch of a missing object will occur. An unexpected
302 missing object will raise an error.
303 --exclude-promisor-objects
305 Omit objects that are known to be in the promisor remote. (This
306 option has the purpose of operating only on locally created
307 objects, so that when we repack, we still maintain a distinction
308 between locally created objects [without .promisor] and objects
309 from the promisor remote [with .promisor].) This is used with
310 partial clone.
311 --keep-unreachable
313 Objects unreachable from the refs in packs named with --unpacked=
314 option are added to the resulting pack, in addition to the
315 reachable objects that are not in packs marked with *.keep files.
316 This implies --revs.
317 --pack-loose-unreachable
319 Pack unreachable loose objects (and their loose counterparts
320 removed). This implies --revs.
321 --unpack-unreachable
323 Keep unreachable objects in loose form. This implies --revs.
324 --delta-islands
326 Restrict delta matches based on "islands". See DELTA ISLANDS below.
327
329 When possible, pack-objects tries to reuse existing on-disk deltas to
330 avoid having to search for new ones on the fly. This is an important
331 optimization for serving fetches, because it means the server can avoid
332 inflating most objects at all and just send the bytes directly from
333 disk. This optimization can’t work when an object is stored as a delta
334 against a base which the receiver does not have (and which we are not
335 already sending). In that case the server "breaks" the delta and has to
336 find a new one, which has a high CPU cost. Therefore it’s important for
337 performance that the set of objects in on-disk delta relationships
338 match what a client would fetch.
339 In a normal repository, this tends to work automatically. The objects
341 are mostly reachable from the branches and tags, and that’s what
342 clients fetch. Any deltas we find on the server are likely to be
343 between objects the client has or will have.
344 But in some repository setups, you may have several related but
346 separate groups of ref tips, with clients tending to fetch those groups
347 independently. For example, imagine that you are hosting several
348 "forks" of a repository in a single shared object store, and letting
349 clients view them as separate repositories through GIT_NAMESPACE or
350 separate repos using the alternates mechanism. A naive repack may find
351 that the optimal delta for an object is against a base that is only
352 found in another fork. But when a client fetches, they will not have
353 the base object, and we’ll have to find a new delta on the fly.
354 A similar situation may exist if you have many refs outside of
356 refs/heads/ and refs/tags/ that point to related objects (e.g.,
357 refs/pull or refs/changes used by some hosting providers). By default,
358 clients fetch only heads and tags, and deltas against objects found
359 only in those other groups cannot be sent as-is.
360 Delta islands solve this problem by allowing you to group your refs
362 into distinct "islands". Pack-objects computes which objects are
363 reachable from which islands, and refuses to make a delta from an
364 object A against a base which is not present in all of A's islands.
365 This results in slightly larger packs (because we miss some delta
366 opportunities), but guarantees that a fetch of one island will not have
367 to recompute deltas on the fly due to crossing island boundaries.
368 When repacking with delta islands the delta window tends to get clogged
370 with candidates that are forbidden by the config. Repacking with a big
371 --window helps (and doesn’t take as long as it otherwise might because
372 we can reject some object pairs based on islands before doing any
373 computation on the content).
374 Islands are configured via the pack.island option, which can be
376 specified multiple times. Each value is a left-anchored regular
377 expressions matching refnames. For example:
378 [pack]
380 island = refs/heads/
381 island = refs/tags/
382 puts heads and tags into an island (whose name is the empty string; see
384 below for more on naming). Any refs which do not match those regular
385 expressions (e.g., refs/pull/123) is not in any island. Any object
386 which is reachable only from refs/pull/ (but not heads or tags) is
387 therefore not a candidate to be used as a base for refs/heads/.
388 Refs are grouped into islands based on their "names", and two regexes
390 that produce the same name are considered to be in the same island. The
391 names are computed from the regexes by concatenating any capture groups
392 from the regex, with a - dash in between. (And if there are no capture
393 groups, then the name is the empty string, as in the above example.)
394 This allows you to create arbitrary numbers of islands. Only up to 14
395 such capture groups are supported though.
396 For example, imagine you store the refs for each fork in
398 refs/virtual/ID, where ID is a numeric identifier. You might then
399 configure:
400 [pack]
402 island = refs/virtual/([0-9]+)/heads/
403 island = refs/virtual/([0-9]+)/tags/
404 island = refs/virtual/([0-9]+)/(pull)/
405 That puts the heads and tags for each fork in their own island (named
407 "1234" or similar), and the pull refs for each go into their own
408 "1234-pull".
409 Note that we pick a single island for each regex to go into, using
411 "last one wins" ordering (which allows repo-specific config to take
412 precedence over user-wide config, and so forth).
413
415 Various configuration variables affect packing, see git-config(1)
416 (search for "pack" and "delta").
417 Notably, delta compression is not used on objects larger than the
419 core.bigFileThreshold configuration variable and on files with the
420 attribute delta set to false.
421
423 git-rev-list(1) git-repack(1) git-prune-packed(1)
424
426 Part of the git(1) suite
427Git 2.43.0 11/20/2023 GIT-PACK-OBJECTS(1)