scripts: eliminate use of find -print0 | xargs -0
[girocco.git] / jobd / gc.sh
blobca90c9b75996204d612a2b4bd463f16bb86d165e
1 #!/bin/sh
3 # NOTE: additional options can be passed to git repack by specifying
4 # them after the project name, for example:
5 # gc.sh my-project -f
7 . @basedir@/shlib.sh
9 set -e
11 if [ $# -lt 1 ]; then
12 echo "Usage: gc.sh projname [extra-repack-args]" >&2
13 exit 1
16 # packing options
17 packopts="--depth=50 --window=50 --window-memory=${var_window_memory:-1g}"
18 quiet=; [ -n "$show_progress" ] || quiet=-q
20 umask 002
21 [ "$cfg_permission_control" != "Hooks" ] || umask 000
22 clean_git_env
24 pidactive() {
25 if _result="$(kill -0 "$1" 2>&1)"; then
26 # process exists and we have permission to signal it
27 return 0
29 case "$_result" in *"not permitted"*)
30 # we do not have permission to signal the process
31 return 0
32 esac
33 # process does not exist
34 return 1
37 createlock() {
38 # A .lock file should only exist for much less than a second.
39 # If we see a stale lock file (> 1h old), remove it and then,
40 # just in case, wait 30 seconds for any process whose .lock
41 # we might have just removed (it's racy) to finish doing what
42 # should take much less than a second to do.
43 _stalelock="$(find "$1.lock" -maxdepth 1 -mmin +60 -print 2>/dev/null)" || :
44 if [ -n "$_stalelock" ]; then
45 rm -f "$_stalelock"
46 sleep 30
48 for _try in p p n; do
49 if (set -C; >"$1.lock") 2>/dev/null; then
50 echo "$1.lock"
51 return 0
53 # delay and try again
54 [ "$_try" != "p" ] || sleep 1
55 done
56 # cannot create lock file
57 return 1
60 # The pre-receive script creates one ref log file per push but we want them to
61 # be coalesced into one ref log file per day. We are guaranteed that any files
62 # we find to coalesce are NOT currently being written to since they are always
63 # written first as temporary files and then moved into place. We attempt to
64 # transfer the most recent modification time to the coalesced log file which
65 # would step on its mod time if it were being written to directly, but if we
66 # find per-process ref log files then it must be a push project and the only
67 # thing that would write directly to the main per-day log file would be a
68 # mirror project so there's actually no conflict.
69 # Also, if the clock is wonky (or was futzed with) we may have both YYYYMMDD
70 # and YYYYMMDD.gz present in which case combine them into YYYYMMDD
71 coalesce_reflogs() {
72 [ -d reflogs ] || return 0
73 rm -f .gc_failed
74 find reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]" -print |
75 while read -r rname; do
76 if [ -e "$rname.gz" ]; then
77 if [ -s "$rname" ]; then
78 # Presumably the .gz file must have been created before the non-gz
79 # file since it had to be uncompressed at some point therefore
80 # we need to append the non-gz contents to it but keep the non-gz
81 # contents timestamp so we rename to YYYYMMDD_ which will sort first
82 # and be picked up in the next step if we are interrupted in the middle.
83 # If a YYYYMMDD_ file already exists we append to it and transfer the
84 # timestamp. Finally we transfer the YYYYMMDD_ timestamp to the result
85 # and remove the YYYYMMDD_ temporary file leaving the result uncompressed.
86 if [ -e "${rname}_" ]; then
87 cat "$rname" >>"${rname}_"
88 touch -r "$rname" "${rname}_"
89 rm -f "$rname"
90 ! [ -e "$rname" ]
91 else
92 mv "$rname" "${rname}_"
94 gzip -d "$rname.gz" </dev/null
95 [ -e "$rname" ] && ! [ -e "$rname.gz" ]
96 cat "${rname}_" >>"$rname"
97 touch -r "${rname}_" "$rname"
98 rm -f "${rname}_"
99 else
100 # Just remove the empty file to resolve the problem
101 rm -f "$rname"
104 done
105 find reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]_*" -print | LC_ALL=C sort |
106 while read -r rname; do
107 logname="${rname%%_*}"
108 # If someone's been futzing with the date, the file we want to
109 # append to could already have been compressed, so we just uncompress
110 # it here. The previous block guarantees we do not have both a compressed
111 # and uncompressed version present at the same time.
112 if [ -e "$logname.gz" ]; then
113 gzip -d "$logname.gz" </dev/null
114 [ -e "$logname" ] && ! [ -e "$logname.gz" ]
116 cat "$rname" >>"$logname"
117 touch -r "$rname" "$logname"
118 rm -f "$rname"
119 if [ -e "$rname" ]; then
120 >.gc_failed
121 echo "! [$proj] failed to remove $rname" >&2
122 exit 1 # will only exit subshell created by "|"
124 done
125 ! [ -e .gc_failed ]
128 # Remove any files in reflogs that are older than $cfg_reflogs_lifetime days
129 prune_reflogs() {
130 [ -d reflogs ] || return 0
131 exp="$(( ${cfg_reflogs_lifetime:-1} * 1440 ))"
132 [ $exp -gt 0 ] || exp=1440
133 [ $exp -le 43200 ] || exp=43200
134 find reflogs -maxdepth 1 -type f -mmin "+$exp" -exec rm -f '{}' + || :
137 # Compact any reflogs that are not today's UTC date unless a .gz version exists
138 compact_reflogs() {
139 [ -d reflogs ] || return 0
140 _td="reflogs/$(TZ=UTC date '+%Y%m%d')"
141 find reflogs -maxdepth 1 -type f -name "[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]" -print |
142 while read -r rname; do
143 [ "$rname" != "$_td" ] || continue
144 ! [ -e "$rname.gz" ] || continue
145 gzip -9 "$rname" </dev/null
146 done
149 # return true if there's more than one objects/pack-<sha>.pack file or
150 # ANY sha-1 files in objects
151 is_dirty() {
152 _packs=$(find objects/pack -type f -name "pack-$octet20.pack" -print | head -n 2 | LC_ALL=C wc -l)
153 if [ $_packs != 1 ] && [ $_packs != 0 ]; then
154 return 0
156 _objs=$(find objects/$octet -type f -name "$octet19" -print 2>/dev/null | head -n 1 | LC_ALL=C wc -l)
157 [ $_objs -ne 0 ]
160 # make sure combine-packs uses the correct Git executable
161 run_combine_packs() {
162 PATH="$var_git_exec_path:$cfg_basedir/bin:$PATH" @basedir@/jobd/combine-packs.sh "$@"
165 # combine the input pack(s) into a new pack (or possibly packs if packSizeLimit set)
166 # input pack names are read from standard input one per line delimited by the first
167 # ':', ' ' or '\n' character on the line (which allows gfi-packs to be read directly)
168 # all arguments, if any, are passed to pack-objects as additional options
169 # returns non-zero on failure AND creates .gc_failed in that case
170 combine_packs() {
171 rm -f .gc_failed
172 find objects/pack -maxdepth 1 -type f -name '*.zap*' -exec rm -f '{}' + || :
173 run_combine_packs --replace "$@" $packopts --all-progress-implied $quiet --non-empty || {
174 >.gc_failed
175 return 1
177 return 0
180 # if the current directory is_gfi_mirror then repack all packs listed in gfi-packs
181 repack_gfi_packs() {
182 [ -n "$gfi_mirror" ] || return 0
183 [ -d objects/pack ] || { rm -f gfi-packs; return 0; }
184 progress "~ [$proj] redeltifying poor quality git fast-import packs"
185 combine_packs --ignore-missing --no-reuse-delta <gfi-packs
186 rm -f gfi-packs
187 return 0
190 # combine small packs into larger pack(s)
191 # we avoid any keep, bndl or bitmap packs
192 # if the optional argument is non-empty even a single small pack will be redeltad
193 combine_small_packs() {
194 _didprogress=
195 _minsmallpacks=2
196 if [ -n "$1" ] && [ -n "$noreusedeltaopt" ]; then
197 _minsmallpacks=1
199 _lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
200 _lpo="$_lpo --object-limit $var_redelta_threshold objects/pack"
201 while
202 _cnt="$(list_packs --count $_lpo)" || :
203 test "${_cnt:-0}" -ge $_minsmallpacks
205 [ -n "$_didprogress" ] || {
206 progress "~ [$proj] combining small packs into a single larger pack"
207 _didprogress=1
209 _newp="$(list_packs $_lpo | combine_packs --names $noreusedeltaopt)"
210 _newc="$(( $(echo "$_newp" | LC_ALL=C wc -w) ))"
211 # be paranoid and exit the loop if we haven't reduced the number of packs
212 [ $_newc -lt $_cnt ] || break
213 _minsmallpacks=2
214 done
215 return 0
218 # Unfortunately git-svn lacks the ability to store newly fetched revisions as a pack.
219 # However, the fetch code conveniently sets .svnpack just before it runs git-svn fetch
220 # so that it's easy to find all the objects that have been fetched by git-svn and
221 # combine them into a pack. The --no-reuse-delta option is meaningless here since
222 # everything to be packed is a loose object and therefore not a delta so deltification
223 # will always take place.
224 make_svn_pack() {
225 [ -f .svnpack ] && [ -n "$svn_mirror" ] || return 0
226 rm -f .svnpackgc
227 mv -f .svnpack .svnpackgc
228 progress "~ [$proj] combining loose git-svn objects into a pack"
229 _newp="$(find objects/$octet -maxdepth 1 -type f -newer .svnpackgc -name "$octet19" -print 2>/dev/null |
230 LC_ALL=C awk -F / '{print $2 $3}' |
231 run_combine_packs --objects --names $packopts --incremental --all-progress-implied $quiet --non-empty)" || {
232 mv -f .svnpackgc .svnpack
233 >.gc_failed
234 return 1
236 if [ -n "$_newp" ]; then
237 # remove the now-redundant loose objects -- this is always safe
238 # even during a concurrent push because a reprepare_packed_git
239 # will be triggered if an object that should be there is not
240 # found thereby finding it in the new pack instead
241 git prune-packed $quiet
243 rm -f .svnpackgc
246 # HEADSHA="$(pack_is_complete /full/path/to/some.pack /full/path/to/packed-refs "$(cat HEAD)")"
247 pack_is_complete() {
248 # Must have a matching .idx file and a non-empty packed-refs file
249 [ -s "${1%.pack}.idx" ] || return 1
250 [ -s "$2" ] || return 1
251 _headsha=
252 case "$3" in
253 $octet20)
254 _headsha="$3"
256 "ref: refs/"?*|"ref:refs/"?*|"refs/"?*)
257 _headmatch="${3#ref:}"
258 _headmatch="${_headmatch# }"
259 _headmatchpat="$(echo "$_headmatch" | LC_ALL=C sed -e 's/\([.$]\)/\\\1/g')"
260 _headsha="$(LC_ALL=C grep -e "^$octet20 $_headmatchpat\$" <"$2" |
261 LC_ALL=C cut -d ' ' -f 1)"
262 case "$_headsha" in $octet20) :;; *)
263 return 1
264 esac
267 # bad HEAD
268 return 1
269 esac
270 rm -rf pack_is_complete_test
271 mkdir pack_is_complete_test
272 mkdir pack_is_complete_test/refs
273 mkdir pack_is_complete_test/objects
274 mkdir pack_is_complete_test/objects/pack
275 echo "$_headsha" >pack_is_complete_test/HEAD
276 ln -s "$1" pack_is_complete_test/objects/pack/
277 ln -s "${1%.pack}.idx" pack_is_complete_test/objects/pack/
278 ln -s "$2" pack_is_complete_test/packed-refs
279 _count="$(git --git-dir=pack_is_complete_test rev-list --count --all 2>/dev/null)" || :
280 rm -rf pack_is_complete_test
281 [ -n "$_count" ] || return 1
282 [ "$_count" -gt 0 ] 2>/dev/null || return 1
283 echo "$_headsha"
286 # On return a "$lockf" will have been created that must be removed when gc is done
287 lock_gc() {
288 # be compatibile with gc.pid file from newer Git releases
289 lockf=gc.pid
290 hn="$(hostname)"
291 active=
292 if [ "$(createlock "$lockf")" ]; then
293 # If $lockf is:
294 # 1) less than 12 hours old
295 # 2) contains two fields (pid hostname) NO trailing NL
296 # 3) the hostname is different OR the pid is still alive
297 # then we exit as another active process is holding the lock
298 if [ "$(find "$lockf" -maxdepth 1 -mmin -720 -print 2>/dev/null)" ]; then
299 apid=
300 ahost=
301 read -r apid ahost ajunk <"$lockf" || :
302 if [ "$apid" ] && [ "$ahost" ]; then
303 if [ "$ahost" != "$hn" ] || pidactive "$apid"; then
304 active=1
308 else
309 echo >&2 "[$proj] unable to create gc.pid.lock file"
310 exit 1
312 if [ -n "$active" ]; then
313 rm -f "$lockf.lock"
314 echo >&2 "[$proj] gc already running on machine '$ahost' pid '$apid'"
315 exit 1
317 printf "%s %s" "$$" "$hn" >"$lockf.lock"
318 chmod 0664 "$lockf.lock"
319 mv -f "$lockf.lock" "$lockf"
322 # Create a repack subdirectory such that running repack in it will pack the
323 # same things that a pack in the normal directory would except that the pack
324 # is guaranteed to be generated in an optimized order by adding a suitable
325 # synthesized ref in the refs/tags namespace (yes, pack-objects.c really does
326 # behave differently depending on the contents of the refs/tags namespace).
327 # Before calling this, pack-refs --all MUST be performed or the wrong pack
328 # will end up being made.
329 # If a ref deletion is pushed after making the repack subdir but before the
330 # the actual repack, the discarded objects will be packed -- no big deal,
331 # they'll get discarded the next time gc runs.
332 # If a fast-forward ref update is pushed after making the repack subdir but
333 # before the actual repack, it will be picked up and the new objects packed
334 # (subject to the normal git repack race about picking such updates up).
335 # If a non-fast-forward ref update is pushed after making the repack subdir but
336 # before the actual repack, it will be picked up like a fast-forward update but
337 # the discarded objects will be included like a ref deletion (until the next
338 # scheduled gc takes place).
339 make_repack_dir() {
340 ! [ -d repack ] || rm -rf repack
341 ! [ -d repack ] || { echo >&2 "[$proj] cannot remove repack subdirectory"; exit 1; }
342 mkdir repack
343 [ -d info ] || mkdir info
344 ln -s ../config repack/config
345 ln -s ../info repack/info
346 ln -s ../objects repack/objects
347 ln -s ../refs repack/refs
348 _lines=$(( $(LC_ALL=C wc -l <packed-refs) ))
349 sed 's, refs/, refs/!/,' <packed-refs >repack/packed-refs
350 optref="$(git rev-list -n 1 --all 2>/dev/null)" || :
351 if [ -n "$optref" ]; then
352 echo "$optref refs/tags/!" >>repack/packed-refs
353 _lines=$(( $_lines + 1 ))
354 echo "$optref" >repack/HEAD
355 else
356 cat HEAD >repack/HEAD
358 if [ $(LC_ALL=C wc -l <repack/packed-refs) -ne "$_lines" ]; then
359 echo >&2 "[$proj] error: make_repack_dir failed packed-refs line count sanity check"
360 exit 1
364 # Remove any crud that's been left behind by interrupted operations
365 # that did not clean up after themselves
366 remove_crud() {
367 # Remove any existing FETCH_HEAD
368 # There can only be a FETCH_HEAD if we've been fetching, not if we've been
369 # receiving pushes (those never create a FETCH_HEAD).
370 # And if we're fetching because we're a mirror, we know we're not fetching right
371 # now since jobd.pl never runs a project's fetch simultaneously with its gc.
372 # Therefore any existing FETCH_HEAD is junk. And it may be many megabytes if
373 # there were a lot of refs.
374 rm -f FETCH_HEAD
376 # remove any existing pack_is_complete_test or repack subdirectories
377 # If either exists when this function is called it's crud
378 rm -rf pack_is_complete_test repack
380 # Remove any stale pack remnants that are more than an hour old.
381 # Stale pack fragments are defined as any pack-<sha1>.ext where .ext is NOT
382 # .pack AND the corresponding .pack DOES NOT exist. A bunch of stale
383 # pack-<sha1>.idx files without their corresponding .pack files are worthless
384 # and just waste space. Normally there shouldn't be any remnants but actually
385 # this can happen when things are interrupted at just the wrong time.
386 # Note that the objects/pack directory is created by git init and should
387 # always exist.
388 find objects/pack -maxdepth 1 -type f -mmin +60 -name "pack-$octet20.?*" -print |
389 LC_ALL=C sed -e 's/^objects\/pack\/pack-//; s/\..*$//' | LC_ALL=C sort -u |
390 while read packsha; do
391 ! [ -e "objects/pack/pack-$packsha.pack" ] || continue
392 rm -f "objects/pack/pack-$packsha".?*
393 done
395 # Remove any stale tmp reflogs files that are more than one hour old.
396 # Since they are created only while the pre-receive hook is running and
397 # all it does is process a bunch of refs passed to it on standard input
398 # it's inconceivable that it would ever take as much as an hour to run.
399 if [ -d reflogs ]; then
400 find reflogs -maxdepth 1 -type f -mmin +60 -name "tmp_*" -exec rm -f '{}' + || :
403 # Remove any stale object tmp_obj_* files that are more than 3 hours old.
404 # Really these files should only exist very briefly so there shouldn't be any
405 # but things happen that can end up leaving them behind.
406 find objects/$octet -maxdepth 1 -type f -mmin +180 -name "tmp_obj_?*" -exec rm -f '{}' + 2>/dev/null || :
408 # Remove any stale pack .keep files that are more than 12 hours old.
409 # We don't do anything to create any permanent pack .keep files, so they must
410 # be remnants from some failed push or something. Removing the .keep will
411 # allow the pack to be properly repacked.
412 find objects/pack -maxdepth 1 -type f -mmin +720 -name "pack-$octet20.keep" -exec rm -f '{}' + || :
414 # Remove any stale tmp_pack_*, tmp_idx_*, tmp_bitmap_*, packtmp-* or .tmp-*-pack* files
415 # that are more than 12 hours old.
416 find objects/pack -maxdepth 1 -type f -mmin +720 \( \
417 -name "tmp_pack_?*" -o -name "tmp_idx_?*" -o -name "tmp_bitmap_?*" -o \
418 -name "packtmp-?*" -o -name ".tmp-?*-pack*" \
419 \) -exec rm -f '{}' + || :
421 # Remove any stale shallow_* files that are more than 12 hours old.
422 # These can be left behind by Git >= 1.8.4.2 and < 2.0.0 when a client
423 # requests a shallow clone.
424 find . -maxdepth 1 -type f -mmin +720 -name "shallow_?*" -exec rm -f '{}' + || :
426 # Remove any stale *.temp files in the objects area that are more than 12 hours old.
427 # This can be stale sha1.temp, or stale *.pack.temp so we kill all stale *.temp.
428 find objects -type f -mmin +720 -name "*.temp" -exec rm -f '{}' + || :
430 # Remove any stale *.lock files in the htmlcache area that might have been left
431 # behind after an abnormal exit during an attempt to update a cached file and
432 # are more than 1 hour old.
433 ! [ -d htmlcache ] || find htmlcache -type f -mmin +60 -name "*.lock" -exec rm -f '{}' + || :
435 # Remove any stale git-svn temp files that are more than 12 hours old.
436 # The git-svn process creates temp files with random 10 character names
437 # in the root of $GIT_DIR. Unfortunately they do not have a recognizable
438 # prefix, so we just have to kill any files with a 10-character name. We
439 # do this only for git-svn mirrors. All characters are chosen from
440 # [A-Za-z0-9_] so we can at least check that and fortunately the only
441 # collision is 'FETCH_HEAD' but that shouldn't matter.
442 # There may also be temp files with a Git_ prefix as well.
443 if [ -n "$svn_mirror" ]; then
444 _randchar='[A-Za-z0-9_]'
445 _randchar2="$_randchar$_randchar"
446 _randchar4="$_randchar2$_randchar2"
447 _randchar10="$_randchar4$_randchar4$_randchar2"
448 find . -maxdepth 1 -type f -mmin +720 -name "$_randchar10" -exec rm -f '{}' + || :
449 find . -maxdepth 1 -type f -mmin +720 -name "Git_*" -exec rm -f '{}' + || :
452 # Remove any stale fast_import_crash_<pid> files that are more than 3 days old.
453 if [ -n "$gfi_mirror" ]; then
454 find . -maxdepth 1 -type f -mmin +4320 -name "fast_import_crash_?*" -exec rm -f '{}' + || :
459 ## Garbage Collection Types
461 ## There are two kinds of possible garbage collection (gc) operations:
463 ## 1. A normal, full gc
464 ## 2. A "mini" gc
466 ## If the full garbage collection interval has expired (or gc has never been
467 ## run), then a normal, full gc will take place. Otherwise, a "mini" gc will
468 ## take place if the file .needsgc exists.
470 ## A "mini" gc is similar to "git gc --auto" in that it may not end up actually
471 ## doing anything unless the right conditions are present so it's not a burden
472 ## to run it often. If the file .needsgc exists, a "mini" gc will occur at
473 ## the next opportunity.
475 ## Note, however, that the .nogc file suppresses ALL gc activity (normal or mini).
478 proj="${1%.git}"
479 shift
480 cd "$cfg_reporoot/$proj.git"
481 [ -d objects/pack ] || { rm -f gfi-packs; mkdir -p objects/pack; }
482 mirror_url="$(get_mirror_url)"
483 svn_mirror=
484 ! is_svn_mirror_url "$mirror_url" || svn_mirror=1
485 gfi_mirror=
486 if [ -f gfi-packs ] && [ -s gfi-packs ] && is_gfi_mirror_url "$mirror_url"; then
487 gfi_mirror=1
490 # If git config --bool --get girocco.redelta is explicitly false then automatic
491 # redelta when there are less than $var_redelta_threshold objects will be suppressed.
492 # On the other hand, if git config --get girocco.redelta is "always" then, on a full
493 # gc only, for the final repack, deltas will always be recomputed.
494 # This can be set on a per-project basis to avoid unusual pathological gc behavior.
495 # Setting this will hurt efficiency of the affected repository.
496 # Note that fast-import packs ALWAYS get new deltas regardless of this setting.
497 noreusedeltaopt="--no-reuse-delta"
498 [ "$(git config --bool --get girocco.redelta 2>/dev/null || :)" != "false" ] || noreusedeltaopt=
499 alwaysredelta=
500 [ "$(git config --get girocco.redelta 2>/dev/null || :)" != "always" ] || alwaysredelta=1
502 trap 'e=$?; rm -f .gc_in_progress; if [ $e != 0 ]; then echo "gc failed dir: $PWD" >&2; fi' EXIT
503 trap 'exit 130' INT
504 trap 'exit 143' TERM
506 # date -R is linux-only, POSIX equivalent is '+%a, %d %b %Y %T %z'
507 datefmt='+%a, %d %b %Y %T %z'
509 isminigc=
510 if check_interval lastgc $cfg_min_gc_interval; then
511 if [ -e .needsgc ]; then
512 isminigc=1
513 else
514 progress "= [$proj] garbage check skip (last at $(config_get lastgc))"
515 exit 0
518 if [ -e .nogc ]; then
519 progress "x [$proj] garbage check disabled"
520 exit 0
523 if [ -n "$isminigc" ]; then
524 # Perform a "mini" gc
525 # Note that .delaygc is ignored here as that's only intended for full gc
526 lock_gc
527 rm -f .allowgc .needsgc
528 remove_crud
529 coalesce_reflogs
530 prune_reflogs
531 compact_reflogs
532 miniactive=
533 if [ -f .svnpack ] && [ -n "$svn_mirror" ]; then
534 miniactive=1
535 progress "+ [$proj] mini garbage check ($(date))"
536 make_svn_pack
538 if [ -z "$cfg_delay_gfi_redelta" ] && [ -n "$gfi_mirror" ]; then
539 # $Girocco::Config::delay_gfi_redelta is false, force redeltification now
540 if [ -z "$miniactive" ]; then
541 miniactive=1
542 progress "+ [$proj] mini garbage check ($(date))"
544 repack_gfi_packs
546 # If there aren't at least 10 non-keep, non-bitmap, non-bndl packs then
547 # don't actually process them yet
548 lpo="--exclude-no-idx --exclude-keep --exclude-bitmap --exclude-bndl --quiet"
549 packcnt="$(list_packs --count $lpo objects/pack)" || :
550 if [ "${packcnt:-0}" -ge 10 ]; then
551 if [ -z "$miniactive" ]; then
552 miniactive=1
553 progress "+ [$proj] mini garbage check ($(date))"
555 if [ -n "$gfi_mirror" ]; then
556 repack_gfi_packs
557 packcnt="$(list_packs --count $lpo objects/pack)" || :
559 # if repack_gfi_packs dropped the pack count to < 10 don't combine
560 if [ "${packcnt:-0}" -ge 10 ]; then
561 combine_small_packs
562 packcnt="$(list_packs --count $lpo objects/pack)" || :
564 # if we still have more than 10 packs trigger a full gc
565 if [ "${packcnt:-0}" -ge 10 ]; then
566 # We shouldn't be in a .delaygc state at this point, but if
567 # we are then nuke it because we really need a full gc now
568 rm -f .delaygc
569 git config --unset gitweb.lastgc
570 rm -f "$lockf"
571 progress "- [$proj] mini garbage check triggering full gc too many packs ($(date))"
572 exit 0
575 rm -f "$lockf"
576 if [ -n "$miniactive" ]; then
577 git update-server-info
578 progress "- [$proj] mini garbage check ($(date))"
579 else
580 progress "= [$proj] mini garbage check nothing but crud removal to do ($(date))"
582 exit 0
585 # Avoid unnecessary garbage collections:
586 # 1. If lastreceive is set and is older than lastgc
587 # -AND-
588 # 2. We are not a fork (! -s alternates) -OR- lastparentgc is older than lastgc
590 # If lastgc is NOT set or lastreceive is NOT set we MUST run gc
591 # If we are a fork and lastparentgc is NOT set we MUST run gc
593 # If the repo is dirty after removing any crud we MUST run gc
595 gcstart="$(date "$datefmt")"
596 skipgc=
597 isfork=
598 ! [ -s objects/info/alternates ] || isfork=1
599 lastparentgcsecs=
600 [ -z "$isfork" ] || lastparentgcsecs="$(config_get_date_seconds lastparentgc)" || :
601 lastreceivesecs=
602 if lastreceivesecs="$(config_get_date_seconds lastreceive)" &&
603 lastgcsecs="$(config_get_date_seconds lastgc)" &&
604 [ $lastreceivesecs -lt $lastgcsecs ]; then
605 # We've run gc since we last received, so maybe we can skip,
606 # check if not fork or fork and lastparentgc < lastgc
607 if [ -n "$isfork" ]; then
608 if [ -n "$lastparentgcsecs" ] &&
609 [ $lastparentgcsecs -lt $lastgcsecs ]; then
610 # We've run gc since our parent ran gc so we can skip
611 skipgc=1
613 else
614 # We don't have any alternates (we're not a forK) so we can skip
615 skipgc=1
619 # Prevent any other simultaneous gc operations
620 lock_gc
622 # At this point, if .allowgc exists, it's now crud to be removed
623 rm -f .allowgc
625 # Ideally we would do this in post-receive, but that would mean duplicating the
626 # logic so it's available in the chroot jail and that's highly undesirable
627 # Instead, since the first gc will be triggered immediately following the first
628 # push, we do the check here as it's quick and harmless if HEAD is already valid
629 check_and_set_head || :
631 # Always get rid of crud
632 remove_crud
634 # Always perform reflogs maintenance
635 coalesce_reflogs
636 prune_reflogs
637 compact_reflogs
639 # Run 'git svn gc' now for svn mirrors
640 if [ -n "$svn_mirror" ]; then
641 git svn gc || :
644 # Skip the actual gc if .delaygc is set
645 if [ -e .delaygc ]; then
646 progress "x [$proj] garbage check delayed (except for crud removal)"
647 rm -f "$lockf"
648 exit 0
651 # Do not skip gc if the repo is dirty
652 if [ -n "$skipgc" ] && ! is_dirty; then
653 progress "= [$proj] garbage check nothing but crud removal to do ($(date))"
654 config_set lastgc "$gcstart"
655 rm -f "$lockf"
656 exit 0
659 bumptime=
660 if [ -n "$isfork" ] && [ -z "$lastparentgcsecs" ]; then
661 # set lastparentgc and then update gcstart to be at least 1 second later
662 config_set lastparentgc "$gcstart"
663 bumptime=1
665 if [ -z "$lastreceivesecs" ]; then
666 # set lastreceive and then update gcstart to be at least 1 second later
667 config_set lastreceive "$gcstart"
668 bumptime=1
670 if [ -n "$bumptime" ]; then
671 sleep 1
672 gcstart="$(date "$datefmt")"
675 progress "+ [$proj] garbage check ($(date))"
677 newdeltas=
678 [ -z "$alwaysredelta" ] || newdeltas=-f
679 if [ -z "$newdeltas" ] && [ -n "$gfi_mirror" ]; then
680 if [ $(list_packs --exclude-no-idx --count objects/pack) -le \
681 $(list_packs --exclude-no-idx --count --quiet --only gfi-packs) ]; then
682 # Don't bother with repack_gfi_packs since everything's being repacked
683 newdeltas=-f
686 if [ -z "$newdeltas" ] && [ -n "$noreusedeltaopt" ] &&
687 [ $(list_packs --exclude-no-idx --count-objects objects/pack) -le $var_redelta_threshold ]; then
688 # There aren't enough objects to worry about so just redelta to get the best pack
689 newdeltas=-f
691 if [ -z "$newdeltas" ]; then
692 # Since we're not going to recompute deltas overall, we need to do the
693 # "mini" maintenance so that we can get more optimal deltas
694 [ -z "$noreusedeltaopt" ] || make_svn_pack
695 repack_gfi_packs
696 force_single_pack_redelta=
697 [ -n "$gfi_mirror" ] || [ -n "$svn_mirror" ] || force_single_pack_redelta=1
698 [ -z "$noreusedeltaopt" ] || combine_small_packs $force_single_pack_redelta
702 ## Safe Pruning In Forks
704 ## We are about to perform garbage collection. We do NOT use the "git gc"
705 ## command directly as it does not provide enough control over the fine details
706 ## that we require. However, we DO maintain a "gc.pid" file during our garbage
707 ## collection so that a simultaneous "git gc" by an administrator will be
708 ## blocked (and similarly we refuse to start garbage collection if we cannot
709 ## create the "gc.pid" file). When we say "gc" in the below description we are
710 ## referring to our "gc.sh" script, NOT the "git gc" command.
712 ## If the project we are running garbage collection (gc) on has any forks we
713 ## must be careful not to remove any objects that while no longer referenced by
714 ## this project (the parent) are still referenced by one or more forks (the
715 ## children) otherwise the children will become corrupt and we can't abide
716 ## corrupt children.
718 ## One way to accomplish this is to simply hard-link all currently existing
719 ## loose objects and packs in the parent into all the children that refer to the
720 ## parent (via a line in their objects/info/alternates file) before beginning
721 ## the gc operation and then relying on a subsequent gc in the child to clean up
722 ## any excess objects/packs. We used to use this strategy but it's very
723 ## inefficient because:
725 ## 1. The disk space used by the old pack(s)/object(s) will not be reclaimed
726 ## until all children (and their children, if any) run gc by which time
727 ## it's quite possible the topmost parent will have run gc again and
728 ## hard-linked yet another old pack down to its children (not to mention
729 ## loose objects).
731 ## 2. As we are now using the "-A" option with "git repack", any new objects
732 ## in the parent that are not referenced by children will continually get
733 ## exploded out of the hard-linked pack in the children whenever the
734 ## children run gc.
736 ## 3. To avoid suboptimal and/or unnecessarily many packs being hard-linked
737 ## into child forks, we must run the "mini" gc maintenance before we
738 ## perform the hard-linking into the children which provides yet another
739 ## source of inefficiency.
741 ## Since we are using the "-A" option to "git repack" (that was not always the
742 ## case) to guarantee we can access old ref values for long enough to send out
743 ## a meaningful mail.sh notification, we now have another, more efficient,
744 ## option available to prevent corruption of child forks that continue to refer
745 ## to objects that are no longer reachable from any ref in the parent.
747 ## The only things that need be copied (or hard-linked) into the child fork(s)
748 ## are those objects that have become unreachable from any ref in the parent.
749 ## They are the only things that could ever be removed by "git prune" and
750 ## therefore the only things we need to prevent the loss of in order to avoid
751 ## corruption of the child fork(s).
753 ## Therefore we now use the following strategy instead to avoid excessive disk
754 ## use and lots of unnecessary loose objects in child forks:
756 ## 1. Run "git repack -A -d -l" in the parent BEFORE doing anything about
757 ## child forks.
759 ## 2. Hard-link all remaining existing loose objects in the parent into the
760 ## immediate child forks.
762 ## 3. Now run "git prune" in the parent.
764 ## With this new strategy we avoid the need to run any "mini" gc maintenance
765 ## before copying (or hard-linking) anything down to the child forks.
766 ## Furthermore, only when the parent performs a non-fast-forward update will
767 ## anything ever be transferred to the children leaving them unperturbed in the
768 ## vast majority of cases. Finally, even if the parent references objects the
769 ## children do not, those objects will no longer continually end up in the
770 ## children as unreachable loose objects after the children run gc.
773 git pack-refs --all
774 make_repack_dir
775 touch .gc_in_progress
776 rm -f .gc_failed bundles/*
777 rm -f objects/pack/pack-*.bndl
778 # We use the -A option with git repack so that unreachable objects can live
779 # on for a time as loose objects. This is particularly helpful if we just
780 # happen to be in the process of sending out a ref update for a ref that was
781 # force updated and the old ref value would have otherwise been removed by
782 # repack because it was now unreachable. Admittedly the window for gc to run
783 # and do that before we manage to send out the ref update is not large, but
784 # it would not be difficult to create such a situation. Unfortunately, when
785 # Git unpacks these unreachable objects it will give them the modification
786 # time of the *.pack file they came out of. This could be very, very old.
787 # If that happens, the subsequent git prune --expire some_time_ago will still
788 # remove the object(s) and our pending ref update will still lose out.
789 # To prevent this from happening and to get the behavior we want, we now
790 # touch the modification time of all pack-<sha>.pack files so that any
791 # loosened objects get a current time. Git does not provide any other
792 # mechanism to do this. We do not want to just touch all loose objects
793 # left after the repack because that would cause objects that were loosened
794 # previously to live on which we definitely do not want.
795 list_packs --exclude-no-idx objects/pack | xargs touch -c 2>/dev/null || :
796 # We wish to keep deltas from our last full pack so if we're not redeltaing
797 # then make sure the .pack associated with the .bitmap has a newer mod time
798 # (If there is no .bitmap then touch the pack with the most objects instead.)
799 if [ -z "$newdeltas" ]; then
800 bmpack="$(list_packs --exclude-no-bitmap --exclude-no-idx --max-matches 1 objects/pack)"
801 [ -n "$bmpack" ] || bmpack="$(list_packs --exclude-no-idx --max-matches 1 --object-limit -1 --include-boundary objects/pack)"
802 if [ -n "$bmpack" ] && [ -f "$bmpack" ] && [ -s "$bmpack" ]; then
803 sleep 1
804 touch -c "$bmpack" 2>/dev/null || :
807 # The git repack command only supports bitmaps if all objects are being packed.
808 # While it is theoretically possible that a project with a non-empty alternates
809 # file ends up packing all objects (because it does not actually use any of the
810 # objects found in the alternates), it's very unlikely. And, in the unlikely
811 # event that did occur, clients would see a message about only using one bitmap
812 # because Git can only use one bitmap at a time and at least one of the
813 # alternates is bound to have a bitmap. Therefore if we see a non-empty
814 # alternates file, we disable writing bitmaps which avoids the warning and any
815 # possibility of a client warning as well.
816 nobm=
817 [ -z "$var_have_git_172" ] || ! [ -s objects/info/alternates ] ||
818 nobm='-c repack.writebitmaps=false -c pack.writebitmaps=false'
819 progress "~ [$proj] running full gc repack${nobm:+ (bitmaps disabled)}"
820 cd repack
821 # We run git repack from the repack subdirectory so we can force optimized packs
822 # to be generated even for repositories that do not have any tagged commits
823 git $nobm repack $packopts -A -d -l $quiet $newdeltas $@
824 cd ..
825 rm -rf repack
826 ! [ -e .gc_failed ] || exit 1
827 # These, if they exist, are now meaningless and need to be removed
828 rm -f gfi-packs .needsgc .svnpack .svnpackgc
829 allpacks="$(echo objects/pack/pack-$octet20.pack)"
830 curhead="$(cat HEAD)"
831 pkrf=
832 ! [ -e packed-refs ] || pkrf=packed-refs
833 eval "reposizek=$(( $(echo 0 $(du -k $pkrf $allpacks 2>/dev/null | LC_ALL=C awk '{print $1}') |
834 LC_ALL=C sed -e 's/ / + /g') ))"
835 git update-server-info
836 # The -A option to `git repack` may have caused some loose objects to pop
837 # out of their packs. We must make these objects group writable so that they
838 # can be freshened by other pushers. Technically we need only do this for
839 # push projects but to enable mirror projects to be more easily converted to
840 # push projects, we go ahead and do it for all projects.
841 { find objects/$octet -type f -name "$octet19" -exec chmod ug+w '{}' + || :; } 2>/dev/null
843 if has_forks "$proj"; then
844 progress "~ [$proj] hard-linking loose objects into immediate child forks"
845 # We have to update the lastparentgc time in the child forks even if they do not get any
846 # new "loose objects" because they need to run gc just in case the parent now has some
847 # objects that used to only be in the child so they can be removed from the child.
848 # For example, a "patch" might be developed first in a fork and then later accepted into
849 # the parent in which case the objects making up the patch in the child fork are now
850 # redundant (since they're now in the parent as well) and need to be removed from the
851 # child fork which can only happen if the child fork runs gc.
852 forkdir="$proj"
853 # It is enough to copy objects just one level down and get_repo_list
854 # takes a regular expression (which is automatically prefixed with '^')
855 # so we can easily match forks exactly one level down from this project
856 get_repo_list "$forkdir/[^/]*:" |
857 while read fork; do
858 # Ignore forks that do not exist or are symbolic links
859 ! [ -L "$cfg_reporoot/$fork.git" ] && [ -d "$cfg_reporoot/$fork.git" ] ||
860 continue
861 # Or do not have a non-zero length alternates file
862 [ -s "$cfg_reporoot/$fork.git/objects/info/alternates" ] ||
863 continue
864 # Match objects in parent project
865 for d in objects/$octet; do
866 [ "$d" != "objects/$octet" ] || continue
867 mkdir -p "$cfg_reporoot/$fork.git/$d"
868 find "$d" -maxdepth 1 -type f -name "$octet19" -exec \
869 "$var_sh_bin" -c 'ln -f "$@" '"'$cfg_reporoot/$fork.git/$d/'" sh '{}' + || :
870 done
871 # Update the fork's lastparentgc date (must be current, not $gcstart)
872 git --git-dir="$cfg_reporoot/$fork.git" config \
873 gitweb.lastparentgc "$(date "$datefmt")"
874 done
877 # The git prune command does not take a -q or --quiet but started outputting
878 # 'Checking connectivity' progress messages in v1.7.9. However, we can
879 # suppress those by piping through cat as it only activates the progress
880 # messages when stderr is a tty. We only expire loose objects older than one
881 # day just in case there's some pending action (such as sending out a ref
882 # update) in progress that might want to examine them. This may leave us with
883 # loose objects. That's okay because at the next gc interval, we will always
884 # run gc if we see any loose objects regardless of whether or not we've seen
885 # any updates or we've received new linked objects from our parent. Note that
886 # in order to keep loose objects that just recently became unreferenced but
887 # have a very old modification date around we rely on some help from both the
888 # update.sh and hooks/pre-receive scripts. Furthermore, since Git v2.2.0
889 # (d3038d22 prune: keep objects reachable from recent objects) an unreachable
890 # object that would otherwise be pruned (because it's too old) will be kept
891 # alive by an unreachable object that refers to it that's not old enough to
892 # be pruned yet.
893 prunecmd='git prune --expire 1_day_ago'
894 [ -n "$show_progress" ] ||
895 prunecmd="{ $prunecmd 2>&1 || touch .gc_failed; } | cat"
896 progress "~ [$proj] pruning expired unreachable loose objects"
897 eval "$prunecmd"
898 ! [ -e .gc_failed ] || exit 1
900 # darcs:// mirrors have a xxx.log file that will grow endlessly
901 # if this is a mirror and the file exists, shorten it to 10000 lines
902 # also take this opportunity to optimize the darcs repo
903 if ! [ -e .nofetch ] && [ -n "$cfg_mirror" ]; then
904 url="$(config_get baseurl)" || :
905 case "$url" in darcs://*)
906 if [ -n "$cfg_mirror_darcs" ]; then
907 url="${url%/}"
908 basedarcs="$(basename "${url#darcs:/}")"
909 if [ -f "$basedarcs.log" ]; then
910 tail -n 10000 "$basedarcs.log" >"$basedarcs.log.$$"
911 mv -f "$basedarcs.log.$$" "$basedarcs.log"
913 if [ -d "$basedarcs.darcs" ]; then
915 cd "$basedarcs.darcs"
916 # without show_progress suppress non-error output
917 [ -n "$show_progress" ] || exec >/dev/null
918 # Note that this does not optimize _darcs/inventories/ :(
919 darcs optimize || :
923 esac
926 # Create a matching .bndl header file for the all-in-one pack we just created
927 # but only if we're not a fork (otherwise the bundle would not be complete)
928 # and we are running at least Git version 1.7.2 (pack_is_complete always fails otherwise)
929 if ! [ -s objects/info/alternates ] && [ -n "$var_have_git_172" ]; then
930 # There should only be one pack in $allpacks but if there was a
931 # simultaneous push...
932 # The one we just created will have a .idx and will NOT have a .keep
933 progress "~ [$proj] creating downloadble bundle header"
934 pkfound=
935 pkhead=
936 for pk in $allpacks; do
937 [ -s "$pk" ] || continue
938 pkbase="${pk%.pack}"
939 [ -s "$pkbase.idx" ] || continue
940 ! [ -e "$pkbase.keep" ] || continue
941 if pkhead="$(pack_is_complete "$PWD/$pk" "$PWD/packed-refs" "$curhead")"; then
942 pkfound="$pkbase"
943 break;
945 done
946 if [ -n "$pkfound" ] && [ -n "$pkhead" ]; then
948 symref=
949 case "$curhead" in "ref: refs/"?*|"ref:refs/"?*|"refs/"?*)
950 symref="${curhead#ref:}"
951 symref="${symref# }"
952 esac
953 bndlurl=
954 [ -z "$cfg_httpbundleurl" ] || bndlurl=" url=$cfg_httpbundleurl/$proj.git/clone.bundle"
955 echo "# v2 git bundle"
956 LC_ALL=C sed -ne "/^$octet20 refs\/[^ $tab]*\$/ p" <packed-refs
957 if [ -n "$symref" ]; then
958 printf "$pkhead HEAD\0symref=HEAD:%s%s\n" "$symref" "$bndlurl"
959 else
960 echo "$pkhead HEAD"
962 echo ""
963 } >"$pkbase.bndl"
964 bndletag="$("$cfg_basedir/bin/rangecgi" --etag -m 1 "$pkbase.bndl" "$pkbase.pack")" || :
965 bndlsha="$(printf '%s' "$bndletag" | git hash-object --stdin)" || :
966 if [ -n "$bndletag" ]; then
967 case "$bndlsha" in $octet20)
968 bndlshatrailer="${bndlsha#????????}"
969 bndlshaprefix="${bndlsha%$bndlshatrailer}"
970 bndlname="$(TZ=UTC date +%Y%m%d_%H%M%S)-${bndlshaprefix:-0}"
971 [ -d bundles ] || mkdir bundles
972 echo "${pkbase#objects/pack/}.bndl" >"bundles/$bndlname"
973 echo "${pkbase#objects/pack/}.pack" >>"bundles/$bndlname"
974 ln -s -f -n "$bndlname" bundles/latest
975 esac
980 # Record the size of this repo as the sum of its *.pack sizes as 1024-byte blocks
981 config_set_raw girocco.reposizek "${reposizek:-0}"
983 # We use $gcstart here to avoid a race where a push occurs during the gc itself
984 # and the next future gc could be incorrectly skipped if we used the current
985 # timestamp here instead
986 config_set lastgc "$gcstart"
987 rm -f "$lockf"
989 progress "- [$proj] garbage check ($(date))"