"deleting unused links" GC phase is too slow

  • Open
  • quality assurance status badge
Details
6 participants
  • John Kehayias
  • Ludovic Courtès
  • Maxim Cournoyer
  • Mark H Weaver
  • Ricardo Wurmus
  • Ricardo Wurmus
Owner
unassigned
Submitted by
Ludovic Courtès
Severity
important
L
L
Ludovic Courtès wrote on 13 Nov 2016 18:41
(address . bug-guix@gnu.org)
87wpg7ffbm.fsf@gnu.org
‘LocalStore::removeUnusedLinks’ traverses all the entries in
/gnu/store/.links and calls lstat(2) on each one of them and checks
‘st_nlink’ to determine whether they can be deleted.

There are two problems: lstat(2) can be slow on spinning disks as found
on hydra.gnu.org, and the algorithm is proportional in the number of
entries in /gnu/store/.links, which is a lot on hydra.gnu.org.

Ludo’.
L
L
Ludovic Courtès wrote on 9 Dec 2016 23:43
(address . 24937@debbugs.gnu.org)(name . Mark H Weaver)(address . mhw@netris.org)
87wpf867v6.fsf@gnu.org
ludo@gnu.org (Ludovic Courtès) skribis:

Toggle quote (8 lines)
> ‘LocalStore::removeUnusedLinks’ traverses all the entries in
> /gnu/store/.links and calls lstat(2) on each one of them and checks
> ‘st_nlink’ to determine whether they can be deleted.
>
> There are two problems: lstat(2) can be slow on spinning disks as found
> on hydra.gnu.org, and the algorithm is proportional in the number of
> entries in /gnu/store/.links, which is a lot on hydra.gnu.org.

On Dec. 2 on guix-sysadmin@gnu.org, Mark described an improvement that
noticeably improved performance:

The idea is to read the entire /gnu/store/.links directory, sort the
entries by inode number, and then iterate over the entries by inode
number, calling 'lstat' on each one and deleting the ones with a link
count of 1.

The reason this is so much faster is because the inodes are stored on
disk in order of inode number, so this leads to a sequential access
pattern on disk instead of a random access pattern.

The difficulty is that the directory is too large to comfortably store
all of the entries in virtual memory. Instead, the entries should be
written to temporary files on disk, and then sorted using merge sort to
ensure sequential access patterns during sorting. Fortunately, this is
exactly what 'sort' does from GNU coreutils.

So, for now, I've implemented this as a pair of small C programs that is
used in a pipeline with GNU sort. The first program simply reads a
directory and writes lines of the form "<inode> <name>" to stdout.
(Unfortunately, "ls -i" calls stat on each entry, so it can't be used).
This is piped through 'sort -n' and then into another small C program
that reads these lines, calls 'lstat' on each one, and deletes the
non-directories with link count 1.

Regarding memory usage, I replied:

Really?

For each entry, we have to store roughly 70 bytes for the file name (or
52 if we consider only the basename), plus 8 bytes for the inode number;
let’s say 64 bytes.

If we have 10 M entries, that’s 700 MB (or 520 MB), which is a lot, but
maybe acceptable?

At worst, we may still see an improvement if we proceed by batches: we
read 10000 directory entries (7 MB), sort them, and stat them, then read
the next 10000 entries. WDYT?

Ludo’.
L
L
Ludovic Courtès wrote on 9 Dec 2016 23:44
control message for bug #24937
(address . control@debbugs.gnu.org)
87vaus67uc.fsf@gnu.org
severity 24937 important
L
L
Ludovic Courtès wrote on 11 Dec 2016 14:46
Re: bug#24937: "deleting unused links" GC phase is too slow
(address . 24937@debbugs.gnu.org)(name . Mark H Weaver)(address . mhw@netris.org)
87lgvm4lzu.fsf@gnu.org
Hello!

Here’s a proposed patch that follows your suggestion, Mark, but places
an upper bound on the number of directory entries loaded in memory.

On my laptop, which has ~500k entries in /gnu/store/.links, the result
is something like this (notice the inode numbers in ‘lstat’ calls):

Toggle snippet (22 lines)
13738 write(4, "gmlo\0\0\0\0\31\0\0\0\0\0\0\0deleting unused "..., 48) = 48
13738 open("/gnu/store/.links", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 8
13738 fstat(8, {st_dev=makedev(8, 2), st_ino=4014083, st_mode=S_IFDIR|0755, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=119224, st_size=60977152, st_atime=2016/12/11-12:01:59, st_mtime=2016/12/11-09:39:45, st_ctime=2016/12/11-09:39:45}) = 0
13738 getdents(8, {{d_ino=4014083, d_off=4294967296, d_reclen=24, d_name="."}
[...]
13738 lstat("/gnu/store/.links/1f2f3g8waxwymp9sl2slcfyara164i8w1y2sz3h9js2fcviv2rnc", {st_dev=makedev(8, 2), st_ino=47, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=88, st_size=41482, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
13738 lstat("/gnu/store/.links/1p25kpyfw354in4kykmgh5sy9h925hnil1jdzgxhz7n6abbws8px", {st_dev=makedev(8, 2), st_ino=65, st_mode=S_IFREG|0444, st_nlink=9, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2313, st_atime=2016/12/08-21:02:26, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/11-01:44:49}) = 0
13738 lstat("/gnu/store/.links/163187g637br9ys5pmshb01wjav53bs1g1a83m7c2alpdyx3yqz2", {st_dev=makedev(8, 2), st_ino=68, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=32, st_size=13561, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
[...]
13738 getdents(8, {{d_ino=4257114, d_off=1734093409898198492,
[...]
13738 lstat("/gnu/store/.links/1m6g06i01ybbkhjjbirjnj7fckw1b772cwygkvbd6v6zgkln7f7m", {st_dev=makedev(8, 2), st_ino=19, st_mode=S_IFREG|0444, st_nlink=4, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2553, st_atime=2016/12/08-21:02:54, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/07-00:05:19}) = 0
13738 lstat("/gnu/store/.links/1ml8n3q55ikn8h60sn67jq1y7z7mvdp5kwr33pqrid2r6kk1d4kb", {st_dev=makedev(8, 2), st_ino=30, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2090, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:21}) = 0
13738 lstat("/gnu/store/.links/1c8cvwlqyyqby3k13cwm40g26pwca5iiz5dcj43xrgn9y91lfvc2", {st_dev=makedev(8, 2), st_ino=33, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=16, st_size=7958, st_atime=2015/11/04-18:55:32, st_mtime=1970/01/01-01:00:01, st_ctime=2016/01/05-11:35:49}) = 0
[...]
13738 getdents(8, {{d_ino=328672,
[...]
13738 lstat("/gnu/store/.links/1l55l59dxb74himmkfzx5v63cv7287i6rjhdns1fdlwajqd73lnz", {st_dev=makedev(8, 2), st_ino=21, st_mode=S_IFREG|0444, st_nlink=31, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=615, st_atime=2016/12/08-21:02:30, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/11-01:44:47}) = 0
13738 lstat("/gnu/store/.links/1c7mm5amw743mb45f1zg4d4r3g549ch35wks9izkcgkx0jirpxsg", {st_dev=makedev(8, 2), st_ino=48, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=360, st_size=176750, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
13738 lstat("/gnu/store/.links/0z5s7b0yk8mfn1np6gk3cdbmpnjgxg1g0l8vfq1aa01zwp06d3f0", {st_dev=makedev(8, 2), st_ino=61, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=1440, st_atime=2016/11/03-20:37:50, st_mtime=1970/01/01-01:00:01, st_ctime=2016/11/07-00:01:57}) = 0

I can’t tell exactly how this affects performance because my machine has
an SSD where this operation takes ~3 seconds on a cold cache. I suspect
it has performance comparable to that of doing all the ‘readdir’ at once
followed by all the ‘lstat’.

Mark, how does that sound?

I’d like to commit it soon if there are no objections.

Thanks,
Ludo’.
Toggle diff (96 lines)
diff --git a/nix/libstore/gc.cc b/nix/libstore/gc.cc
index 72eff52..db58603 100644
--- a/nix/libstore/gc.cc
+++ b/nix/libstore/gc.cc
@@ -545,6 +545,9 @@ void LocalStore::tryToDelete(GCState & state, const Path & path)
}
+/* Like 'dirent', but with just what we need. */
+typedef std::pair<Path, ino_t> MiniDirEntry;
+
/* Unlink all files in /nix/store/.links that have a link count of 1,
which indicates that there are no other links and so they can be
safely deleted. FIXME: race condition with optimisePath(): we
@@ -555,32 +558,57 @@ void LocalStore::removeUnusedLinks(const GCState & state)
AutoCloseDir dir = opendir(linksDir.c_str());
if (!dir) throw SysError(format("opening directory `%1%'") % linksDir);
+ /* Maximum number of entries stored in memory; each 'MiniDirEntry' takes
+ ~80 bytes. */
+ const size_t maxEntries = 100000;
+
long long actualSize = 0, unsharedSize = 0;
- struct dirent * dirent;
- while (errno = 0, dirent = readdir(dir)) {
- checkInterrupt();
- string name = dirent->d_name;
- if (name == "." || name == "..") continue;
- Path path = linksDir + "/" + name;
-
- struct stat st;
- if (lstat(path.c_str(), &st) == -1)
- throw SysError(format("statting `%1%'") % path);
-
- if (st.st_nlink != 1) {
- unsigned long long size = st.st_blocks * 512ULL;
- actualSize += size;
- unsharedSize += (st.st_nlink - 1) * size;
- continue;
- }
-
- printMsg(lvlTalkative, format("deleting unused link `%1%'") % path);
-
- if (unlink(path.c_str()) == -1)
- throw SysError(format("deleting `%1%'") % path);
-
- state.results.bytesFreed += st.st_blocks * 512;
+ bool endOfDir = false;
+
+ while (!endOfDir) {
+ /* Read as many entries as possible at once, up to 'maxEntries'. */
+ std::list<MiniDirEntry> entries;
+ struct dirent * dirent;
+ while (errno = 0,
+ (entries.size() < maxEntries) && (dirent = readdir(dir))) {
+ checkInterrupt();
+ string name = dirent->d_name;
+ if (name == "." || name == "..") continue;
+ entries.emplace_back(MiniDirEntry(name, dirent->d_ino));
+ }
+
+ endOfDir = (dirent == NULL);
+
+ /* Sort entries by inode number to minimize disk seeks induced by the
+ following 'lstat' calls. */
+ entries.sort([] (const MiniDirEntry& e1, const MiniDirEntry& e2) {
+ return e1.second < e2.second;
+ });
+
+ for (auto && item: entries) {
+ checkInterrupt();
+
+ Path path = linksDir + "/" + item.first;
+
+ struct stat st;
+ if (lstat(path.c_str(), &st) == -1)
+ throw SysError(format("statting `%1%'") % path);
+
+ if (st.st_nlink != 1) {
+ unsigned long long size = st.st_blocks * 512ULL;
+ actualSize += size;
+ unsharedSize += (st.st_nlink - 1) * size;
+ continue;
+ }
+
x+ printMsg(lvlTalkative, format("deleting unused link `%1%'") % path);
+
+ if (unlink(path.c_str()) == -1)
+ throw SysError(format("deleting `%1%'") % path);
+
+ state.results.bytesFreed += st.st_blocks * 512;
+ }
}
struct stat st;
M
M
Mark H Weaver wrote on 11 Dec 2016 15:23
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
87twaaa6j9.fsf@netris.org
ludo@gnu.org (Ludovic Courtès) writes:

Toggle quote (34 lines)
> Here’s a proposed patch that follows your suggestion, Mark, but places
> an upper bound on the number of directory entries loaded in memory.
>
> On my laptop, which has ~500k entries in /gnu/store/.links, the result
> is something like this (notice the inode numbers in ‘lstat’ calls):
>
> 13738 write(4, "gmlo\0\0\0\0\31\0\0\0\0\0\0\0deleting unused "..., 48) = 48
> 13738 open("/gnu/store/.links", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 8
> 13738 fstat(8, {st_dev=makedev(8, 2), st_ino=4014083, st_mode=S_IFDIR|0755, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=119224, st_size=60977152, st_atime=2016/12/11-12:01:59, st_mtime=2016/12/11-09:39:45, st_ctime=2016/12/11-09:39:45}) = 0
> 13738 getdents(8, {{d_ino=4014083, d_off=4294967296, d_reclen=24, d_name="."}
> [...]
> 13738 lstat("/gnu/store/.links/1f2f3g8waxwymp9sl2slcfyara164i8w1y2sz3h9js2fcviv2rnc", {st_dev=makedev(8, 2), st_ino=47, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=88, st_size=41482, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
> 13738 lstat("/gnu/store/.links/1p25kpyfw354in4kykmgh5sy9h925hnil1jdzgxhz7n6abbws8px", {st_dev=makedev(8, 2), st_ino=65, st_mode=S_IFREG|0444, st_nlink=9, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2313, st_atime=2016/12/08-21:02:26, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/11-01:44:49}) = 0
> 13738 lstat("/gnu/store/.links/163187g637br9ys5pmshb01wjav53bs1g1a83m7c2alpdyx3yqz2", {st_dev=makedev(8, 2), st_ino=68, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=32, st_size=13561, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
> [...]
> 13738 getdents(8, {{d_ino=4257114, d_off=1734093409898198492,
> [...]
> 13738 lstat("/gnu/store/.links/1m6g06i01ybbkhjjbirjnj7fckw1b772cwygkvbd6v6zgkln7f7m", {st_dev=makedev(8, 2), st_ino=19, st_mode=S_IFREG|0444, st_nlink=4, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2553, st_atime=2016/12/08-21:02:54, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/07-00:05:19}) = 0
> 13738 lstat("/gnu/store/.links/1ml8n3q55ikn8h60sn67jq1y7z7mvdp5kwr33pqrid2r6kk1d4kb", {st_dev=makedev(8, 2), st_ino=30, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2090, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:21}) = 0
> 13738 lstat("/gnu/store/.links/1c8cvwlqyyqby3k13cwm40g26pwca5iiz5dcj43xrgn9y91lfvc2", {st_dev=makedev(8, 2), st_ino=33, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=16, st_size=7958, st_atime=2015/11/04-18:55:32, st_mtime=1970/01/01-01:00:01, st_ctime=2016/01/05-11:35:49}) = 0
> [...]
> 13738 getdents(8, {{d_ino=328672,
> [...]
> 13738 lstat("/gnu/store/.links/1l55l59dxb74himmkfzx5v63cv7287i6rjhdns1fdlwajqd73lnz", {st_dev=makedev(8, 2), st_ino=21, st_mode=S_IFREG|0444, st_nlink=31, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=615, st_atime=2016/12/08-21:02:30, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/11-01:44:47}) = 0
> 13738 lstat("/gnu/store/.links/1c7mm5amw743mb45f1zg4d4r3g549ch35wks9izkcgkx0jirpxsg", {st_dev=makedev(8, 2), st_ino=48, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=360, st_size=176750, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
> 13738 lstat("/gnu/store/.links/0z5s7b0yk8mfn1np6gk3cdbmpnjgxg1g0l8vfq1aa01zwp06d3f0", {st_dev=makedev(8, 2), st_ino=61, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=1440, st_atime=2016/11/03-20:37:50, st_mtime=1970/01/01-01:00:01, st_ctime=2016/11/07-00:01:57}) = 0
>
> I can’t tell exactly how this affects performance because my machine has
> an SSD where this operation takes ~3 seconds on a cold cache. I suspect
> it has performance comparable to that of doing all the ‘readdir’ at once
> followed by all the ‘lstat’.
>
> Mark, how does that sound?

I think we should sort the entire directory using merge sort backed to
disk files. If we load chunks of the directory, sort them and process
them individually, I expect that this will increase the amount of I/O
required by a non-trivial factor. In each pass, we would load blocks of
inodes from disk, almost all of which are likely to be present in the
store and thus linked from the directory, but in this scheme we will
process only a small number of them and drop the rest on the floor to be
read again in the next pass. Given that even my fairly optimal
implementation takes about 35 minutes to run on Hydra, I'd prefer to
avoid multiplying that by a non-trivial factor.

Why not just use GNU sort? It already exists, and does exactly what we
need.

If you object to using an external program for some reason, I would
prefer to re-implement a similar algorithm in the daemon.

Thanks,
Mark
L
L
Ludovic Courtès wrote on 11 Dec 2016 19:02
(name . Mark H Weaver)(address . mhw@netris.org)(address . 24937@debbugs.gnu.org)
87twaa2vjx.fsf@gnu.org
Mark H Weaver <mhw@netris.org> skribis:

Toggle quote (47 lines)
> ludo@gnu.org (Ludovic Courtès) writes:
>
>> Here’s a proposed patch that follows your suggestion, Mark, but places
>> an upper bound on the number of directory entries loaded in memory.
>>
>> On my laptop, which has ~500k entries in /gnu/store/.links, the result
>> is something like this (notice the inode numbers in ‘lstat’ calls):
>>
>> 13738 write(4, "gmlo\0\0\0\0\31\0\0\0\0\0\0\0deleting unused "..., 48) = 48
>> 13738 open("/gnu/store/.links", O_RDONLY|O_NONBLOCK|O_DIRECTORY|O_CLOEXEC) = 8
>> 13738 fstat(8, {st_dev=makedev(8, 2), st_ino=4014083, st_mode=S_IFDIR|0755, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=119224, st_size=60977152, st_atime=2016/12/11-12:01:59, st_mtime=2016/12/11-09:39:45, st_ctime=2016/12/11-09:39:45}) = 0
>> 13738 getdents(8, {{d_ino=4014083, d_off=4294967296, d_reclen=24, d_name="."}
>> [...]
>> 13738 lstat("/gnu/store/.links/1f2f3g8waxwymp9sl2slcfyara164i8w1y2sz3h9js2fcviv2rnc", {st_dev=makedev(8, 2), st_ino=47, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=88, st_size=41482, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
>> 13738 lstat("/gnu/store/.links/1p25kpyfw354in4kykmgh5sy9h925hnil1jdzgxhz7n6abbws8px", {st_dev=makedev(8, 2), st_ino=65, st_mode=S_IFREG|0444, st_nlink=9, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2313, st_atime=2016/12/08-21:02:26, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/11-01:44:49}) = 0
>> 13738 lstat("/gnu/store/.links/163187g637br9ys5pmshb01wjav53bs1g1a83m7c2alpdyx3yqz2", {st_dev=makedev(8, 2), st_ino=68, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=32, st_size=13561, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
>> [...]
>> 13738 getdents(8, {{d_ino=4257114, d_off=1734093409898198492,
>> [...]
>> 13738 lstat("/gnu/store/.links/1m6g06i01ybbkhjjbirjnj7fckw1b772cwygkvbd6v6zgkln7f7m", {st_dev=makedev(8, 2), st_ino=19, st_mode=S_IFREG|0444, st_nlink=4, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2553, st_atime=2016/12/08-21:02:54, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/07-00:05:19}) = 0
>> 13738 lstat("/gnu/store/.links/1ml8n3q55ikn8h60sn67jq1y7z7mvdp5kwr33pqrid2r6kk1d4kb", {st_dev=makedev(8, 2), st_ino=30, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=2090, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:21}) = 0
>> 13738 lstat("/gnu/store/.links/1c8cvwlqyyqby3k13cwm40g26pwca5iiz5dcj43xrgn9y91lfvc2", {st_dev=makedev(8, 2), st_ino=33, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=16, st_size=7958, st_atime=2015/11/04-18:55:32, st_mtime=1970/01/01-01:00:01, st_ctime=2016/01/05-11:35:49}) = 0
>> [...]
>> 13738 getdents(8, {{d_ino=328672,
>> [...]
>> 13738 lstat("/gnu/store/.links/1l55l59dxb74himmkfzx5v63cv7287i6rjhdns1fdlwajqd73lnz", {st_dev=makedev(8, 2), st_ino=21, st_mode=S_IFREG|0444, st_nlink=31, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=615, st_atime=2016/12/08-21:02:30, st_mtime=1970/01/01-01:00:01, st_ctime=2016/12/11-01:44:47}) = 0
>> 13738 lstat("/gnu/store/.links/1c7mm5amw743mb45f1zg4d4r3g549ch35wks9izkcgkx0jirpxsg", {st_dev=makedev(8, 2), st_ino=48, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=360, st_size=176750, st_atime=2015/03/10-11:29:06, st_mtime=1970/01/01-01:00:01, st_ctime=2015/11/25-11:29:20}) = 0
>> 13738 lstat("/gnu/store/.links/0z5s7b0yk8mfn1np6gk3cdbmpnjgxg1g0l8vfq1aa01zwp06d3f0", {st_dev=makedev(8, 2), st_ino=61, st_mode=S_IFREG|0444, st_nlink=2, st_uid=0, st_gid=0, st_blksize=4096, st_blocks=8, st_size=1440, st_atime=2016/11/03-20:37:50, st_mtime=1970/01/01-01:00:01, st_ctime=2016/11/07-00:01:57}) = 0
>>
>> I can’t tell exactly how this affects performance because my machine has
>> an SSD where this operation takes ~3 seconds on a cold cache. I suspect
>> it has performance comparable to that of doing all the ‘readdir’ at once
>> followed by all the ‘lstat’.
>>
>> Mark, how does that sound?
>
> I think we should sort the entire directory using merge sort backed to
> disk files. If we load chunks of the directory, sort them and process
> them individually, I expect that this will increase the amount of I/O
> required by a non-trivial factor. In each pass, we would load blocks of
> inodes from disk, almost all of which are likely to be present in the
> store and thus linked from the directory, but in this scheme we will
> process only a small number of them and drop the rest on the floor to be
> read again in the next pass. Given that even my fairly optimal
> implementation takes about 35 minutes to run on Hydra, I'd prefer to
> avoid multiplying that by a non-trivial factor.

Sure, though it’s not obvious to me how much of a difference it makes;
my guess is that processing in large chunks is already a win, but we’d
have to measure.

Toggle quote (3 lines)
> Why not just use GNU sort? It already exists, and does exactly what we
> need.

Does ‘sort’ manage to avoid reading whole files in memory? If not, I
think it wouldn’t buy us anything to use it.

Toggle quote (3 lines)
> If you object to using an external program for some reason, I would
> prefer to re-implement a similar algorithm in the daemon.

Yeah, I’d rather avoid serializing the list of file names/inode number
pairs just to invoke ‘sort’ on that.

Also, what algorithm are you referring to?

Thanks for the quick feedback!

Ludo’.
M
M
Mark H Weaver wrote on 11 Dec 2016 20:27
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
87lgvm9sgq.fsf@netris.org
ludo@gnu.org (Ludovic Courtès) writes:

Toggle quote (17 lines)
> Mark H Weaver <mhw@netris.org> skribis:
>
>> I think we should sort the entire directory using merge sort backed to
>> disk files. If we load chunks of the directory, sort them and process
>> them individually, I expect that this will increase the amount of I/O
>> required by a non-trivial factor. In each pass, we would load blocks of
>> inodes from disk, almost all of which are likely to be present in the
>> store and thus linked from the directory, but in this scheme we will
>> process only a small number of them and drop the rest on the floor to be
>> read again in the next pass. Given that even my fairly optimal
>> implementation takes about 35 minutes to run on Hydra, I'd prefer to
>> avoid multiplying that by a non-trivial factor.
>
> Sure, though it’s not obvious to me how much of a difference it makes;
> my guess is that processing in large chunks is already a win, but we’d
> have to measure.

I agree, it would surely be a win. Given that it currently takes on the
order of a day to run this phase on Hydra, if your proposed method takes
2 hours, that would be a huge win, but still not good, IMO. Even 35
minutes is slower than I'd like.

Toggle quote (5 lines)
>> Why not just use GNU sort? It already exists, and does exactly what we
>> need.
>
> Does ‘sort’ manage to avoid reading whole files in memory?

Yes, it does. I monitored the 'sort' process when I first ran my
optimized pipeline. It created about 10 files in /tmp, approximately 70
megabytes each as I recall, and then read them all concurrently while
writing the sorted output.

My guess is that it reads a manageable chunk of the input, sorts it in
memory, and writes it to a temporary file. I guess it repeats this
process, writing multiple temporary files, until the entire input is
consumed, and then reads all of those temporary files, merging them
together into the output stream.

Toggle quote (6 lines)
>> If you object to using an external program for some reason, I would
>> prefer to re-implement a similar algorithm in the daemon.
>
> Yeah, I’d rather avoid serializing the list of file names/inode number
> pairs just to invoke ‘sort’ on that.

Sure, I agree that it would be better to avoid that, but IMO not at the
cost of using O(N) memory instead of O(1) memory, nor at the cost of
multiplying the amount of disk I/O by a non-trivial factor.

Toggle quote (2 lines)
> Also, what algorithm are you referring to?

The algorithm I described above, which I guess is close to what GNU sort
does.

Thanks,
Mark
L
L
Ludovic Courtès wrote on 13 Dec 2016 01:00
(name . Mark H Weaver)(address . mhw@netris.org)(address . 24937@debbugs.gnu.org)
87d1gwvgu0.fsf@gnu.org
Mark H Weaver <mhw@netris.org> skribis:

Toggle quote (24 lines)
> ludo@gnu.org (Ludovic Courtès) writes:
>
>> Mark H Weaver <mhw@netris.org> skribis:
>>
>>> I think we should sort the entire directory using merge sort backed to
>>> disk files. If we load chunks of the directory, sort them and process
>>> them individually, I expect that this will increase the amount of I/O
>>> required by a non-trivial factor. In each pass, we would load blocks of
>>> inodes from disk, almost all of which are likely to be present in the
>>> store and thus linked from the directory, but in this scheme we will
>>> process only a small number of them and drop the rest on the floor to be
>>> read again in the next pass. Given that even my fairly optimal
>>> implementation takes about 35 minutes to run on Hydra, I'd prefer to
>>> avoid multiplying that by a non-trivial factor.
>>
>> Sure, though it’s not obvious to me how much of a difference it makes;
>> my guess is that processing in large chunks is already a win, but we’d
>> have to measure.
>
> I agree, it would surely be a win. Given that it currently takes on the
> order of a day to run this phase on Hydra, if your proposed method takes
> 2 hours, that would be a huge win, but still not good, IMO. Even 35
> minutes is slower than I'd like.

Of course.

I did some measurements with the attached program on chapters, which is
a Xen VM with spinning disks underneath, similar to hydra.gnu.org. It
has 600k entries in /gnu/store/.links.

Here’s a comparison of the “optimal” mode (bulk stats after we’ve
fetched all the dirents) vs. the “semi-interleaved” mode (doing bulk
stats every 100,000 dirents):

Toggle snippet (19 lines)
ludo@guix:~$ gcc -std=gnu99 -Wall links-traversal.c -DMODE=3
ludo@guix:~$ sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
ludo@guix:~$ time ./a.out
603858 dir_entries, 157 seconds
stat took 1 seconds

real 2m38.508s
user 0m0.324s
sys 0m1.824s
ludo@guix:~$ gcc -std=gnu99 -Wall links-traversal.c -DMODE=2
ludo@guix:~$ sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
ludo@guix:~$ time ./a.out
3852 dir_entries, 172 seconds (including stat)

real 2m51.827s
user 0m0.312s
sys 0m1.808s

Semi-interleaved is ~12% slower here (not sure how reproducible that is
though).

Toggle quote (16 lines)
>>> Why not just use GNU sort? It already exists, and does exactly what we
>>> need.
>>
>> Does ‘sort’ manage to avoid reading whole files in memory?
>
> Yes, it does. I monitored the 'sort' process when I first ran my
> optimized pipeline. It created about 10 files in /tmp, approximately 70
> megabytes each as I recall, and then read them all concurrently while
> writing the sorted output.
>
> My guess is that it reads a manageable chunk of the input, sorts it in
> memory, and writes it to a temporary file. I guess it repeats this
> process, writing multiple temporary files, until the entire input is
> consumed, and then reads all of those temporary files, merging them
> together into the output stream.

OK. That seems to be that the comment above ‘sortlines’ in sort.c
describes.

Toggle quote (10 lines)
>>> If you object to using an external program for some reason, I would
>>> prefer to re-implement a similar algorithm in the daemon.
>>
>> Yeah, I’d rather avoid serializing the list of file names/inode number
>> pairs just to invoke ‘sort’ on that.
>
> Sure, I agree that it would be better to avoid that, but IMO not at the
> cost of using O(N) memory instead of O(1) memory, nor at the cost of
> multiplying the amount of disk I/O by a non-trivial factor.

Understood.

sort.c in Coreutils is very big, and we surely don’t want to duplicate
all that. Yet, I’d rather not shell out to ‘sort’.

Do you know how many entries are in .links on hydra.gnu.org? If it
performs comparably to chapters, the timings suggests it should have
around 10.5M entries.

Thanks!

Ludo’.
#include <unistd.h>
#include <dirent.h>
#include <sys/types.h>
#include <stdlib.h>
#include <stdio.h>
#include <sys/time.h>
#include <string.h>
#include <sys/stat.h>
#include <assert.h>

#define STAT_INTERLEAVED 1
#define STAT_SEMI_INTERLEAVED 2
#define STAT_OPTIMAL 3

struct entry
{
char *name;
ino_t inode;
};

#define MAX_ENTRIES 1000000
static struct entry dir_entries[MAX_ENTRIES];

int
main ()
{
struct timeval start, end;

/* For useful timings, do:
sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches' */
gettimeofday (&start, NULL);
DIR *links = opendir ("/gnu/store/.links");

size_t count = 0;

#if MODE != STAT_INTERLEAVED
void sort_entries (void)
{
int entry_lower (const void *a, const void *b)
{
return ((struct entry *)a)->inode < ((struct entry *)b)->inode;
}

qsort (dir_entries, count, sizeof (struct entry),
entry_lower);
}
#endif

void stat_entries (void)
{
for (size_t i = 0; i < count; i++)
{
struct stat st;
lstat (dir_entries[i].name, &st);
}
}

for (struct dirent *entry = readdir (links);
entry != NULL;
entry = readdir (links))
{
assert (count < MAX_ENTRIES);
dir_entries[count].name = strdup (entry->d_name);
dir_entries[count].inode = entry->d_ino;
#if MODE == STAT_INTERLEAVED
struct stat st;
lstat (entry->d_name, &st);
#endif

#if MODE == STAT_SEMI_INTERLEAVED
if (count++ >= 100000)
{
sort_entries ();
stat_entries ();
count = 0;
}
#else
count++;
#endif
}

#if MODE == STAT_SEMI_INTERLEAVED
sort_entries ();
stat_entries ();
#endif

gettimeofday (&end, NULL);
printf ("%zi dir_entries, %zi seconds"
#if MODE != STAT_OPTIMAL
" (including stat)"
#endif
"\n", count,
end.tv_sec - start.tv_sec);

#if MODE == STAT_OPTIMAL
sort_entries ();
gettimeofday (&start, NULL);
stat_entries ();
gettimeofday (&end, NULL);

printf ("stat took %zi seconds\n", end.tv_sec - start.tv_sec);
#endif

return EXIT_SUCCESS;
}
M
M
Mark H Weaver wrote on 13 Dec 2016 05:09
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
8737hs1nd1.fsf@netris.org
Do as you wish. I don't have time to continue discussing this.

Mark
M
M
Mark H Weaver wrote on 13 Dec 2016 13:48
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
87wpf4yoz0.fsf@netris.org
ludo@gnu.org (Ludovic Courtès) writes:

Toggle quote (4 lines)
> I did some measurements with the attached program on chapters, which is
> a Xen VM with spinning disks underneath, similar to hydra.gnu.org. It
> has 600k entries in /gnu/store/.links.

I just want to point out that 600k inodes use 150 megabytes of disk
space on ext4, which is small enough to fit in the cache, so the disk
I/O will not be multiplied for such a small test case.

Toggle quote (25 lines)
> Here’s a comparison of the “optimal” mode (bulk stats after we’ve
> fetched all the dirents) vs. the “semi-interleaved” mode (doing bulk
> stats every 100,000 dirents):
>
> ludo@guix:~$ gcc -std=gnu99 -Wall links-traversal.c -DMODE=3
> ludo@guix:~$ sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
> ludo@guix:~$ time ./a.out
> 603858 dir_entries, 157 seconds
> stat took 1 seconds
>
> real 2m38.508s
> user 0m0.324s
> sys 0m1.824s
> ludo@guix:~$ gcc -std=gnu99 -Wall links-traversal.c -DMODE=2
> ludo@guix:~$ sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
> ludo@guix:~$ time ./a.out
> 3852 dir_entries, 172 seconds (including stat)
>
> real 2m51.827s
> user 0m0.312s
> sys 0m1.808s
>
> Semi-interleaved is ~12% slower here (not sure how reproducible that is
> though).

This directory you're testing on is more than an order of magnitude
smaller than Hydra's when it's full. Unlike in your test above, all of
the inodes in Hydra's store won't fit in the cache.

In my opinion, the reason Hydra performs so poorly is because efficiency
and scalability are apparently very low priorities in the design of the
software running on it. Unfortunately, I feel that my advice in this
area is discarded more often than not.

Toggle quote (19 lines)
>>>> Why not just use GNU sort? It already exists, and does exactly what we
>>>> need.
>>>
>>> Does ‘sort’ manage to avoid reading whole files in memory?
>>
>> Yes, it does. I monitored the 'sort' process when I first ran my
>> optimized pipeline. It created about 10 files in /tmp, approximately 70
>> megabytes each as I recall, and then read them all concurrently while
>> writing the sorted output.
>>
>> My guess is that it reads a manageable chunk of the input, sorts it in
>> memory, and writes it to a temporary file. I guess it repeats this
>> process, writing multiple temporary files, until the entire input is
>> consumed, and then reads all of those temporary files, merging them
>> together into the output stream.
>
> OK. That seems to be that the comment above ‘sortlines’ in sort.c
> describes.

well-studied problem with a long history.

Toggle quote (6 lines)
>>>> If you object to using an external program for some reason, I would
>>>> prefer to re-implement a similar algorithm in the daemon.
>>>
>>> Yeah, I’d rather avoid serializing the list of file names/inode number
>>> pairs just to invoke ‘sort’ on that.

I'm fairly sure that the overhead of serializing the file names and
inode numbers is *far* less than the overhead you would add by iterating
over the inodes in multiple passes.

Toggle quote (9 lines)
>> Sure, I agree that it would be better to avoid that, but IMO not at the
>> cost of using O(N) memory instead of O(1) memory, nor at the cost of
>> multiplying the amount of disk I/O by a non-trivial factor.
>
> Understood.
>
> sort.c in Coreutils is very big, and we surely don’t want to duplicate
> all that. Yet, I’d rather not shell out to ‘sort’.

The "shell" would not be involved here at all, just the "sort" program.
I guess you dislike launching external processes? Can you explain why?

Guix-daemon launches external processes for building derivations, so why
is using one for garbage collection a problem? Emacs, a program that
you cite in your talks as having many qualities that we seek to emulate,
does not shy away from using external programs.

Toggle quote (2 lines)
> Do you know how many entries are in .links on hydra.gnu.org?

"df -i /gnu" indicates that it currently has about 5.5M inodes, but
that's with only 29% of the disk in use. A few days ago, when the disk
was full, assuming that the average file size is the same, it may have
had closer to 5.5M / 0.29 ~= 19M inodes, which is over 30 times as many
as used in your measurements above. On ext4, which uses 256-byte
inodes, that's about 5 gigabytes of inodes.

Mark
L
L
Ludovic Courtès wrote on 13 Dec 2016 18:02
(name . Mark H Weaver)(address . mhw@netris.org)
87fulrsqxx.fsf@gnu.org
Hello Mark,

Mark H Weaver <mhw@netris.org> skribis:

Toggle quote (10 lines)
> ludo@gnu.org (Ludovic Courtès) writes:
>
>> I did some measurements with the attached program on chapters, which is
>> a Xen VM with spinning disks underneath, similar to hydra.gnu.org. It
>> has 600k entries in /gnu/store/.links.
>
> I just want to point out that 600k inodes use 150 megabytes of disk
> space on ext4, which is small enough to fit in the cache, so the disk
> I/O will not be multiplied for such a small test case.

Right. That’s the only spinning-disk machine I could access without
problem. :-/

Ricardo, Roel: would you be able to run that links-traversal.c from
on a machine with a big store, as described at

Toggle quote (7 lines)
>> Semi-interleaved is ~12% slower here (not sure how reproducible that is
>> though).
>
> This directory you're testing on is more than an order of magnitude
> smaller than Hydra's when it's full. Unlike in your test above, all of
> the inodes in Hydra's store won't fit in the cache.

Good point. I’m trying my best to get performance figures, there’s no
doubt we could do better!

Toggle quote (5 lines)
> In my opinion, the reason Hydra performs so poorly is because efficiency
> and scalability are apparently very low priorities in the design of the
> software running on it. Unfortunately, I feel that my advice in this
> area is discarded more often than not.

Well, as you know, I’m currently traveling, yet I take the time to
answer your email at night; I think this should suggest that far from
discarding your advice, I very much value it.

I’m a maintainer though, so I’m trying to understand the problem better.
It’s not just about finding the “optimal” solution, but also about
finding a tradeoff between the benefits and the maintainability costs.

Toggle quote (6 lines)
>> sort.c in Coreutils is very big, and we surely don’t want to duplicate
>> all that. Yet, I’d rather not shell out to ‘sort’.
>
> The "shell" would not be involved here at all, just the "sort" program.
> I guess you dislike launching external processes? Can you explain why?

I find that passing strings around among programs is inelegant
(subjective), but I don’t think you’re really looking to argue about
that, are you? :-)

It remains that, if invoking ‘sort’ appears to be preferable *both* from
performance and maintenance viewpoints, then it’s a good choice. That
may be the case, but again, I prefer to have figures to back that.

Toggle quote (7 lines)
>> Do you know how many entries are in .links on hydra.gnu.org?
>
> "df -i /gnu" indicates that it currently has about 5.5M inodes, but
> that's with only 29% of the disk in use. A few days ago, when the disk
> was full, assuming that the average file size is the same, it may have
> had closer to 5.5M / 0.29 ~= 19M inodes,

OK, good to know.

Thanks!

Ludo’.
R
R
Ricardo Wurmus wrote on 13 Dec 2016 18:18
(name . Ludovic Courtès)(address . ludo@gnu.org)
87vaunbvcu.fsf@mdc-berlin.de
Ludovic Courtès <ludo@gnu.org> writes:

Toggle quote (5 lines)
> Ricardo, Roel: would you be able to run that links-traversal.c from
> <https://debbugs.gnu.org/cgi/bugreport.cgi?filename=links-traversal.c;bug=24937;msg=25;att=1>
> on a machine with a big store, as described at
> <https://debbugs.gnu.org/cgi/bugreport.cgi?bug=24937#25>?

I just ran this on my workstation in the office where I regularly build
packages. Here’s the output of “df -i /gnu”

Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/mapper/fedora-root 3301376 1098852 2202524 34% /

Probably not large enough to derive conclusions about hydra’s behaviour.

[I can’t run it on the shared store at the MDC because NFS performance is
too poor. I recently ran “guix gc --optimize” to dedupe the shared
store (post-build deduplication is disabled since a few weeks) and it’s
at 3,197,489 used inodes.]

Here are the results of running the link-traversal code on my
workstation:

Toggle snippet (21 lines)
rwurmus in ~: gcc -std=gnu99 -Wall links-traversal.c -DMODE=3
rwurmus in ~: sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
rwurmus in ~: time ./a.out
412825 dir_entries, 107 seconds
stat took 0 seconds

real 1m47.264s
user 0m0.214s
sys 0m1.314s

rwurmus in ~: gcc -std=gnu99 -Wall links-traversal.c -DMODE=2
rwurmus in ~: sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
rwurmus in ~: time ./a.out
12821 dir_entries, 107 seconds (including stat)

real 1m46.475s
user 0m0.201s
sys 0m1.309s


--
Ricardo
M
M
Mark H Weaver wrote on 15 Dec 2016 02:19
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
87h9663s6e.fsf@netris.org
I apologize for losing my patience earlier.

Mark
R
R
Ricardo Wurmus wrote on 16 Apr 2020 15:26
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
87ftd3muhp.fsf@elephly.net
Ricardo Wurmus <ricardo.wurmus@mdc-berlin.de> writes:

Toggle quote (44 lines)
> Ludovic Courtès <ludo@gnu.org> writes:
>
>> Ricardo, Roel: would you be able to run that links-traversal.c from
>> <https://debbugs.gnu.org/cgi/bugreport.cgi?filename=links-traversal.c;bug=24937;msg=25;att=1>
>> on a machine with a big store, as described at
>> <https://debbugs.gnu.org/cgi/bugreport.cgi?bug=24937#25>?
>
> I just ran this on my workstation in the office where I regularly build
> packages. Here’s the output of “df -i /gnu”
>
> Filesystem Inodes IUsed IFree IUse% Mounted on
> /dev/mapper/fedora-root 3301376 1098852 2202524 34% /
>
> Probably not large enough to derive conclusions about hydra’s behaviour.
>
> [I can’t run it on the shared store at the MDC because NFS performance is
> too poor. I recently ran “guix gc --optimize” to dedupe the shared
> store (post-build deduplication is disabled since a few weeks) and it’s
> at 3,197,489 used inodes.]
>
> Here are the results of running the link-traversal code on my
> workstation:
>
> --8<---------------cut here---------------start------------->8---
> rwurmus in ~: gcc -std=gnu99 -Wall links-traversal.c -DMODE=3
> rwurmus in ~: sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
> rwurmus in ~: time ./a.out
> 412825 dir_entries, 107 seconds
> stat took 0 seconds
>
> real 1m47.264s
> user 0m0.214s
> sys 0m1.314s
>
> rwurmus in ~: gcc -std=gnu99 -Wall links-traversal.c -DMODE=2
> rwurmus in ~: sudo sh -c 'echo 3 > /proc/sys/vm/drop_caches'
> rwurmus in ~: time ./a.out
> 12821 dir_entries, 107 seconds (including stat)
>
> real 1m46.475s
> user 0m0.201s
> sys 0m1.309s
> --8<---------------cut here---------------end--------------->8---

I ran this for the first time on ci.guix.gnu.org, which has a very big
store (currently at around 29TB).

df -i /gnu:

Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/sdb1 610021376 132350406 477670970 22% /gnu

I had to increase the number of MAX_ENTRIES to 135000000.

I forgot to drop caches initially. This is the first run:

Toggle snippet (10 lines)
root@berlin ~ [env]# gcc links-traversal.c -DMODE=3 -o links-traversal
root@berlin ~ [env]# time ./links-traversal
57079502 dir_entries, 3906 seconds
stat took 136 seconds

real 67m48.145s
user 0m59.575s
sys 2m30.065s

I aborted the run after I dropped caches after 67 minutes.

I’m going to continue testing on one of the build nodes, and I’ll try
using statx.

--
Ricardo
R
R
Ricardo Wurmus wrote on 16 Apr 2020 16:27
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
87eesnmrow.fsf@elephly.net
Here are more benchmarks on one of the build nodes. It doesn’t nearly
have as many used inodes as ci.guix.gnu.org, but I could fill it up if
necessary.

root@hydra-guix-127 ~# df -i /gnu/
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/sda3 28950528 2796829 26153699 10% /

root@hydra-guix-127 ~# ls -1 /gnu/store/.links | wc -l
2017395

I tested all three modes with statx and with lstat. The
links-traversal-statx.c is attached below.

* mode 1 + statx

Toggle snippet (27 lines)
root@hydra-guix-127 ~ [env]# gcc -Wall -std=c99 links-traversal-statx.c -DMODE=1 -D_GNU_SOURCE=1 -o links-traversal
links-traversal-statx.c:53:8: warning: ?stat_entries? defined but not used [-Wunused-function]
53 | void stat_entries (void)
| ^~~~~~~~~~~~
root@hydra-guix-127 ~ [env]# echo 3 > /proc/sys/vm/drop_caches
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 9 seconds (including stat)

real 0m9.176s
user 0m0.801s
sys 0m4.236s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 4 seconds (including stat)

real 0m3.556s
user 0m0.708s
sys 0m2.848s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 4 seconds (including stat)

real 0m3.553s
user 0m0.599s
sys 0m2.954s
root@hydra-guix-127 ~ [env]#


* mode 2 + statx

Toggle snippet (24 lines)
root@hydra-guix-127 ~ [env]# gcc -Wall -std=c99 links-traversal-statx.c -DMODE=2 -D_GNU_SOURCE=1 -o links-traversal
root@hydra-guix-127 ~ [env]# echo 3 > /proc/sys/vm/drop_caches
root@hydra-guix-127 ~ [env]# time ./links-traversal
17377 dir_entries, 10 seconds (including stat)

real 0m9.598s
user 0m1.210s
sys 0m4.257s
root@hydra-guix-127 ~ [env]# time ./links-traversal
17377 dir_entries, 4 seconds (including stat)

real 0m4.094s
user 0m0.988s
sys 0m3.107s
root@hydra-guix-127 ~ [env]# time ./links-traversal
17377 dir_entries, 4 seconds (including stat)

real 0m4.095s
user 0m0.933s
sys 0m3.162s
root@hydra-guix-127 ~ [env]#


* mode 3 + statx

Toggle snippet (26 lines)
root@hydra-guix-127 ~ [env]# gcc -Wall -std=c99 links-traversal-statx.c -DMODE=3 -D_GNU_SOURCE=1 -o links-traversal^C
root@hydra-guix-127 ~ [env]# echo 3 > /proc/sys/vm/drop_caches
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 7 seconds
stat took 3 seconds

real 0m9.992s
user 0m1.411s
sys 0m4.221s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 1 seconds
stat took 2 seconds

real 0m4.265s
user 0m1.120s
sys 0m3.145s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 2 seconds
stat took 2 seconds

real 0m4.267s
user 0m1.072s
sys 0m3.195s
root@hydra-guix-127 ~ [env]#

Now with just lstat:

* mode 1 + lstat

Toggle snippet (26 lines)
root@hydra-guix-127 ~ [env]# gcc -Wall -std=c99 links-traversal.c -DMODE=1 -D_GNU_SOURCE=1 -o links-traversal
links-traversal.c:49:8: warning: ?stat_entries? defined but not used [-Wunused-function]
49 | void stat_entries (void)
| ^~~~~~~~~~~~
root@hydra-guix-127 ~ [env]# echo 3 > /proc/sys/vm/drop_caches
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 9 seconds (including stat)

real 0m9.303s
user 0m0.748s
sys 0m4.397s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 4 seconds (including stat)

real 0m3.526s
user 0m0.540s
sys 0m2.987s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 3 seconds (including stat)

real 0m3.519s
user 0m0.600s
sys 0m2.919s
root@hydra-guix-127 ~ [env]#

* mode 2 + lstat

Toggle snippet (23 lines)
root@hydra-guix-127 ~ [env]# gcc -Wall -std=c99 links-traversal.c -DMODE=2 -D_GNU_SOURCE=1 -o links-traversal
root@hydra-guix-127 ~ [env]# echo 3 > /proc/sys/vm/drop_caches
root@hydra-guix-127 ~ [env]# time ./links-traversal
17377 dir_entries, 9 seconds (including stat)

real 0m9.614s
user 0m1.205s
sys 0m4.250s
root@hydra-guix-127 ~ [env]# time ./links-traversal
17377 dir_entries, 4 seconds (including stat)

real 0m4.060s
user 0m1.052s
sys 0m3.008s
root@hydra-guix-127 ~ [env]# time ./links-traversal
17377 dir_entries, 4 seconds (including stat)

real 0m4.057s
user 0m0.984s
sys 0m3.073s
root@hydra-guix-127 ~ [env]#

* mode 3 + lstat

Toggle snippet (26 lines)
root@hydra-guix-127 ~ [env]# gcc -Wall -std=c99 links-traversal.c -DMODE=3 -D_GNU_SOURCE=1 -o links-traversal
root@hydra-guix-127 ~ [env]# echo 3 > /proc/sys/vm/drop_caches
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 6 seconds
stat took 3 seconds

real 0m9.767s
user 0m1.270s
sys 0m4.339s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 2 seconds
stat took 2 seconds

real 0m4.234s
user 0m1.136s
sys 0m3.097s
root@hydra-guix-127 ~ [env]# time ./links-traversal
2017397 dir_entries, 1 seconds
stat took 2 seconds

real 0m4.222s
user 0m1.052s
sys 0m3.170s
root@hydra-guix-127 ~ [env]#

They are all very close, so I think I need to work with a bigger store
to see a difference.

Or perhaps I did something silly because I don’t know C… If so please
let me know.

--
Ricardo
L
L
Ludovic Courtès wrote on 17 Apr 2020 10:16
(name . Ricardo Wurmus)(address . rekado@elephly.net)(address . 24937@debbugs.gnu.org)
874ktieddi.fsf@gnu.org
Hi Ricardo,

Thanks for running this benchmark!

Ricardo Wurmus <rekado@elephly.net> skribis:

Toggle quote (3 lines)
> root@hydra-guix-127 ~# ls -1 /gnu/store/.links | wc -l
> 2017395

That’s not a lot, my laptop has 2.8M links.

It’s interesting to see that system time remains at ~4.2s in all modes.
So the only thing that modes 2 and 3 achieve is increasing CPU time.
It’s as if the order in which files are stat’d had no impact on I/O
performance.

Ludo’.
R
R
Ricardo Wurmus wrote on 17 Apr 2020 10:28
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 24937@debbugs.gnu.org)
874ktims89.fsf@elephly.net
Ludovic Courtès <ludo@gnu.org> writes:

Toggle quote (5 lines)
>> root@hydra-guix-127 ~# ls -1 /gnu/store/.links | wc -l
>> 2017395
>
> That’s not a lot, my laptop has 2.8M links.

Let me rerun this after copying a few thousand store items from
ci.guix.gnu.org over. Maybe we’ll see the different times diverge then.

--
Ricardo
L
L
Ludovic Courtès wrote on 9 Nov 2021 15:44
(address . 24937@debbugs.gnu.org)(name . Maxim Cournoyer)(address . maxim.cournoyer@gmail.com)
87pmr9l76m.fsf@gnu.org
Hi!

ludo@gnu.org (Ludovic Courtès) skribis:

Toggle quote (8 lines)
> ‘LocalStore::removeUnusedLinks’ traverses all the entries in
> /gnu/store/.links and calls lstat(2) on each one of them and checks
> ‘st_nlink’ to determine whether they can be deleted.
>
> There are two problems: lstat(2) can be slow on spinning disks as found
> on hydra.gnu.org, and the algorithm is proportional in the number of
> entries in /gnu/store/.links, which is a lot on hydra.gnu.org.

Taking a step back, we could perhaps mitigate this with heuristics to
reduce the number of entries in .links:

1. Do not deduplicate files with a size lower than some threshold;

2. Delete links with st_nlink <= 3 (instead of <= 2); that would
prevent *further* deduplication of those files, but they’d already
have two instances sharing the same inode;

3. Stop deduplicating once the number of entries in .links has reached
a certain threshold.

For #1, a key insight is that about 30% of the files actually
deduplicated (in my store, where /gnu/store/.links has 2.2M entries) are
smaller than 1 KiB:
… but 85% of them have at most 4 links (thus, saving up to 2 KiB by
deduplicating):
Attachment: nlink-small.png
On my laptop, we’re talking about space savings of 325 MiB, a tiny
fraction of my store:

Toggle snippet (6 lines)
scheme@(guile-user)> (saved-space (filter (lambda (file)
(< (deduplicated-file-size file) 1024))
l))
$40 = 325914739

Files smaller than 1 KiB represent 35% of the entries in .links:
Attachment: size.png
By not deduplicating files smaller than 1 KiB, we’d reduce the number of
entries by 35%, which should already have a tangible impact on
performance. It’d be a “mitigation” more than a “fix”, but it has a
good work/reward ratio.

We could conduct a similar analysis for #2.

#3 is more difficult to implement because you cannot know the number of
entries in .links until you’ve traversed it (note that currently
deduplication stops when link(2) returns ENOSPC in .links).

I’m attaching the script I’ve used for that, derived from an earlier
experiment¹. You’re welcome to give it a spin!

Thoughts?

Ludo’.

L
L
Ludovic Courtès wrote on 9 Nov 2021 16:00
(address . 24937@debbugs.gnu.org)(name . Maxim Cournoyer)(address . maxim.cournoyer@gmail.com)
87fss5l6gm.fsf@gnu.org
Ludovic Courtès <ludo@gnu.org> skribis:

Toggle quote (8 lines)
> On my laptop, we’re talking about space savings of 325 MiB, a tiny
> fraction of my store:
>
> scheme@(guile-user)> (saved-space (filter (lambda (file)
> (< (deduplicated-file-size file) 1024))
> l))
> $40 = 325914739

For files < 4 KiB, the savings are ~2 GiB, roughly 1% of my store.

Ludo’.
M
L
L
Ludovic Courtès wrote on 13 Nov 2021 17:56
(name . Maxim Cournoyer)(address . maxim.cournoyer@gmail.com)(address . 24937@debbugs.gnu.org)
87v90wat9n.fsf@gnu.org
Hi,

Maxim Cournoyer <maxim.cournoyer@gmail.com> skribis:

Toggle quote (3 lines)
> I haven't done any analysis, just grabbed the result, but here it what
> it looks for me:

There’s a bit more than 35% of deduplicated files that are < 1KiB, and
not much to be gained by deduplicating them.

On IRC several people shared the results on their machine; several had
similar results, and one person had a lot more of those small files (50%
of deduplicated files were < 1KiB).

The chart (with a kinda bogus layout) below is perhaps more interesting:
it shows the contribution of files below a certain size to the overall
space savings.
In a nutshell:

• Files < 1KiB contribute to 0.3% of the space savings;

• Files < 4KiB contribute to 2.5% of the space savings;

• Files < 256KiB contribute to 42% of the space savings.

You can create this plot with:

Toggle snippet (22 lines)
(make-scatter-plot #:title "Contribution to space savings"
#:write-to-png "/tmp/space-saving-contribution.png"
#:chart-width 1000
#:y-axis-label "contribution (%)"
#:x-axis-label "size (B)"
#:log-x-base 2
#:min-x 513
#:data
(let ((total (saved-space l)))
`(("contribution"
,@(map (lambda (size)
(cons size
(/ (saved-space (filter (lambda (file)
(< (deduplicated-file-size
file)
size))
l))
total .01)))
(map (cut expt 2 <>)
(iota 12 10 1)))))))

You can also compute individual points like this:

Toggle snippet (17 lines)
scheme@(guile-user)> (/ (saved-space (filter (lambda (file)
(< (deduplicated-file-size file) 1024))
l))
(saved-space l) 1.)
$60 = 0.0034284626558736746
scheme@(guile-user)> (/ (saved-space (filter (lambda (file)
(< (deduplicated-file-size file) 4096))
l))
(saved-space l) 1.)
$62 = 0.025190871178467848
scheme@(guile-user)> (/ (saved-space (filter (lambda (file)
(< (deduplicated-file-size file) (expt 2 18)))
l))
(saved-space l) 1.)
$65 = 0.42411104869782185

Choosing a deduplication threshold of 2KiB or 4KiB would have a
negligible impact on disk usage on my machine.

Thanks,
Ludo’.
L
L
Ludovic Courtès wrote on 13 Nov 2021 22:37
[PATCH 1/2] =?UTF-8?q?tests:=20Factorize=20'file=3D=3F'.?=
(address . 24937@debbugs.gnu.org)(name . Ludovic Courtès)(address . ludo@gnu.org)
20211113213745.2601-1-ludo@gnu.org
* guix/tests.scm (file=?): Add optional 'stat' parameter. Add fast
patch comparing inode numbers.
* tests/gexp.scm ("imported-files with file-like objects"): Remove
'file=?' procedure and use the one from (guix tests).
---
guix/tests.scm | 30 +++++++++++++++++-------------
tests/gexp.scm | 11 +++--------
2 files changed, 20 insertions(+), 21 deletions(-)

Toggle diff (81 lines)
diff --git a/guix/tests.scm b/guix/tests.scm
index fc3d521163..e1c194340c 100644
--- a/guix/tests.scm
+++ b/guix/tests.scm
@@ -1,5 +1,5 @@
;;; GNU Guix --- Functional package management for GNU
-;;; Copyright © 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020 Ludovic Courtès <ludo@gnu.org>
+;;; Copyright © 2013-2021 Ludovic Courtès <ludo@gnu.org>
;;;
;;; This file is part of GNU Guix.
;;;
@@ -182,18 +182,22 @@ (define (random-bytevector n)
(loop (1+ i)))
bv))))
-(define (file=? a b)
- "Return true if files A and B have the same type and same content."
- (and (eq? (stat:type (lstat a)) (stat:type (lstat b)))
- (case (stat:type (lstat a))
- ((regular)
- (equal?
- (call-with-input-file a get-bytevector-all)
- (call-with-input-file b get-bytevector-all)))
- ((symlink)
- (string=? (readlink a) (readlink b)))
- (else
- (error "what?" (lstat a))))))
+(define* (file=? a b #:optional (stat lstat))
+ "Return true if files A and B have the same type and same content. Call
+STAT to obtain file metadata."
+ (let ((sta (stat a)) (stb (stat b)))
+ (and (eq? (stat:type sta) (stat:type stb))
+ (case (stat:type sta)
+ ((regular)
+ (or (and (= (stat:ino sta) (stat:ino stb))
+ (= (stat:dev sta) (stat:dev stb)))
+ (equal?
+ (call-with-input-file a get-bytevector-all)
+ (call-with-input-file b get-bytevector-all))))
+ ((symlink)
+ (string=? (readlink a) (readlink b)))
+ (else
+ (error "what?" (stat a)))))))
(define (canonical-file? file)
"Return #t if FILE is in the store, is read-only, and its mtime is 1."
diff --git a/tests/gexp.scm b/tests/gexp.scm
index 39a47d4e8c..0758a49f5f 100644
--- a/tests/gexp.scm
+++ b/tests/gexp.scm
@@ -1,5 +1,5 @@
;;; GNU Guix --- Functional package management for GNU
-;;; Copyright © 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021 Ludovic Courtès <ludo@gnu.org>
+;;; Copyright © 2014-2021 Ludovic Courtès <ludo@gnu.org>
;;; Copyright © 2021 Maxime Devos <maximedevos@telenet.be>
;;;
;;; This file is part of GNU Guix.
@@ -827,19 +827,14 @@ (define (canonical-file? file)
(files -> `(("a/b/c" . ,q-scm)
("p/q" . ,plain)))
(drv (imported-files files)))
- (define (file=? file1 file2)
- ;; Assume deduplication is in place.
- (= (stat:ino (stat file1))
- (stat:ino (stat file2))))
-
(mbegin %store-monad
(built-derivations (list (pk 'drv drv)))
(mlet %store-monad ((dir -> (derivation->output-path drv))
(plain* (text-file "foo" "bar!"))
(q-scm* (interned-file q-scm "c")))
(return
- (and (file=? (string-append dir "/a/b/c") q-scm*)
- (file=? (string-append dir "/p/q") plain*)))))))
+ (and (file=? (string-append dir "/a/b/c") q-scm* stat)
+ (file=? (string-append dir "/p/q") plain* stat)))))))
(test-equal "gexp-modules & ungexp"
'((bar) (foo))
--
2.33.0
L
L
Ludovic Courtès wrote on 13 Nov 2021 22:37
[PATCH 2/2] daemon: Do not deduplicate files smaller than 4 KiB.
(address . 24937@debbugs.gnu.org)(name . Ludovic Courtès)(address . ludo@gnu.org)
20211113213745.2601-2-ludo@gnu.org
Files smaller than 4 KiB typically represent ~60% of the entries in
/gnu/store/.links but only contribute to ~2.5% of the space savings
afforded by deduplication.

Not considering these files for deduplication speeds up file insertion
in the store and, more importantly, leaves 'removeUnusedLinks' with
fewer entries to traverse, thereby speeding it up proportionally.


* config-daemon.ac: Remove symlink hard link check and CAN_LINK_SYMLINK
definition.
* guix/store/deduplication.scm (%deduplication-minimum-size): New
variable.
(deduplicate)[loop]: Do not recurse when FILE's size is below
%DEDUPLICATION-MINIMUM-SIZE.
(dump-port): New procedure.
(dump-file/deduplicate)[hash]: Turn into...
[dump-and-compute-hash]: ... this thunk.
Call 'deduplicate' only when SIZE is greater than
%DEDUPLICATION-MINIMUM-SIZE; otherwise call 'dump-port'.
* nix/libstore/gc.cc (LocalStore::removeUnusedLinks): Drop files where
st.st_size < deduplicationMinSize.
* nix/libstore/local-store.hh (deduplicationMinSize): New declaration.
* nix/libstore/optimise-store.cc (deduplicationMinSize): New variable.
(LocalStore::optimisePath_): Return when PATH is a symlink or smaller
than 'deduplicationMinSize'.
* tests/derivations.scm ("identical files are deduplicated"): Produce
files bigger than %DEDUPLICATION-MINIMUM-SIZE.
* tests/nar.scm ("restore-file-set with directories (signed, valid)"):
Likewise.
* tests/store-deduplication.scm ("deduplicate, below %deduplication-minimum-size"):
New test.
("deduplicate", "deduplicate, ENOSPC"): Produce files bigger than
%DEDUPLICATION-MINIMUM-SIZE.
* tests/store.scm ("substitute, deduplication"): Likewise.
---
config-daemon.ac | 11 -------
guix/store/deduplication.scm | 57 ++++++++++++++++++++++++++++------
nix/libstore/gc.cc | 4 ++-
nix/libstore/local-store.hh | 3 ++
nix/libstore/optimise-store.cc | 15 +++++----
tests/derivations.scm | 14 ++++++---
tests/nar.scm | 7 +++--
tests/store-deduplication.scm | 41 ++++++++++++++++++++----
tests/store.scm | 4 ++-
9 files changed, 114 insertions(+), 42 deletions(-)

Toggle diff (363 lines)
diff --git a/config-daemon.ac b/config-daemon.ac
index 5ddc740600..86306effe1 100644
--- a/config-daemon.ac
+++ b/config-daemon.ac
@@ -94,17 +94,6 @@ if test "x$guix_build_daemon" = "xyes"; then
AC_CHECK_FUNCS([lutimes lchown posix_fallocate sched_setaffinity \
statvfs nanosleep strsignal statx])
- dnl Check whether the store optimiser can optimise symlinks.
- AC_MSG_CHECKING([whether it is possible to create a link to a symlink])
- ln -s bla tmp_link
- if ln tmp_link tmp_link2 2> /dev/null; then
- AC_MSG_RESULT(yes)
- AC_DEFINE(CAN_LINK_SYMLINK, 1, [Whether link() works on symlinks.])
- else
- AC_MSG_RESULT(no)
- fi
- rm -f tmp_link tmp_link2
-
dnl Check for <locale>.
AC_LANG_PUSH(C++)
AC_CHECK_HEADERS([locale])
diff --git a/guix/store/deduplication.scm b/guix/store/deduplication.scm
index cd9660174c..8a59adad39 100644
--- a/guix/store/deduplication.scm
+++ b/guix/store/deduplication.scm
@@ -1,6 +1,6 @@
;;; GNU Guix --- Functional package management for GNU
;;; Copyright © 2017 Caleb Ristvedt <caleb.ristvedt@cune.org>
-;;; Copyright © 2018, 2019, 2020 Ludovic Courtès <ludo@gnu.org>
+;;; Copyright © 2018-2021 Ludovic Courtès <ludo@gnu.org>
;;;
;;; This file is part of GNU Guix.
;;;
@@ -22,12 +22,13 @@
(define-module (guix store deduplication)
#:use-module (gcrypt hash)
- #:use-module (guix build utils)
+ #:use-module ((guix build utils) #:hide (dump-port))
#:use-module (guix build syscalls)
#:use-module (guix base32)
#:use-module (srfi srfi-11)
#:use-module (srfi srfi-34)
#:use-module (srfi srfi-35)
+ #:use-module (rnrs bytevectors)
#:use-module (rnrs io ports)
#:use-module (ice-9 ftw)
#:use-module (ice-9 match)
@@ -37,6 +38,31 @@ (define-module (guix store deduplication)
dump-file/deduplicate
copy-file/deduplicate))
+;; TODO: Remove once 'dump-port' in (guix build utils) has an optional 'len'
+;; parameter.
+(define* (dump-port in out
+ #:optional len
+ #:key (buffer-size 16384))
+ "Read LEN bytes from IN (or as much as possible if LEN is #f) and write it
+to OUT, using chunks of BUFFER-SIZE bytes."
+ (define buffer
+ (make-bytevector buffer-size))
+
+ (let loop ((total 0)
+ (bytes (get-bytevector-n! in buffer 0
+ (if len
+ (min len buffer-size)
+ buffer-size))))
+ (or (eof-object? bytes)
+ (and len (= total len))
+ (let ((total (+ total bytes)))
+ (put-bytevector out buffer 0 bytes)
+ (loop total
+ (get-bytevector-n! in buffer 0
+ (if len
+ (min (- len total) buffer-size)
+ buffer-size)))))))
+
(define (nar-sha256 file)
"Gives the sha256 hash of a file and the size of the file in nar form."
(let-values (((port get-hash) (open-sha256-port)))
@@ -127,6 +153,12 @@ (define temp-link
(unless (= EMLINK (system-error-errno args))
(apply throw args)))))))
+(define %deduplication-minimum-size
+ ;; Size below which files are not deduplicated. This avoids adding too many
+ ;; entries to '.links', which would slow down 'removeUnusedLinks' while
+ ;; saving little space. Keep in sync with optimize-store.cc.
+ 4096)
+
(define* (deduplicate path hash #:key (store (%store-directory)))
"Check if a store item with sha256 hash HASH already exists. If so,
replace PATH with a hardlink to the already-existing one. If not, register
@@ -144,13 +176,16 @@ (define links-directory
((file . properties)
(unless (member file '("." ".."))
(let* ((file (string-append path "/" file))
+ (st (lstat file))
(type (match (assoc-ref properties 'type)
((or 'unknown #f)
- (stat:type (lstat file)))
+ (stat:type st))
(type type))))
- (loop file type
- (and (not (eq? 'directory type))
- (nar-sha256 file)))))))
+ (unless (< (stat:size st)
+ %deduplication-minimum-size)
+ (loop file type
+ (and (not (eq? 'directory type))
+ (nar-sha256 file))))))))
(scandir* path))
(let ((link-file (string-append links-directory "/"
(bytevector->nix-base32-string hash))))
@@ -222,9 +257,9 @@ (define* (dump-file/deduplicate file input size type
This procedure is suitable as a #:dump-file argument to 'restore-file'. When
used that way, it deduplicates files on the fly as they are restored, thereby
-removing the need to a deduplication pass that would re-read all the files
+removing the need for a deduplication pass that would re-read all the files
down the road."
- (define hash
+ (define (dump-and-compute-hash)
(call-with-output-file file
(lambda (output)
(let-values (((hash-port get-hash)
@@ -236,7 +271,11 @@ (define hash
(close-port hash-port)
(get-hash)))))
- (deduplicate file hash #:store store))
+ (if (>= size %deduplication-minimum-size)
+ (deduplicate file (dump-and-compute-hash) #:store store)
+ (call-with-output-file file
+ (lambda (output)
+ (dump-port input output size)))))
(define* (copy-file/deduplicate source target
#:key (store (%store-directory)))
diff --git a/nix/libstore/gc.cc b/nix/libstore/gc.cc
index e1d0765154..16519116e4 100644
--- a/nix/libstore/gc.cc
+++ b/nix/libstore/gc.cc
@@ -606,7 +606,9 @@ void LocalStore::removeUnusedLinks(const GCState & state)
throw SysError(format("statting `%1%'") % path);
#endif
- if (st.st_nlink != 1) {
+ /* Drop links for files smaller than 'deduplicationMinSize', even if
+ they have more than one hard link. */
+ if (st.st_nlink != 1 && st.st_size >= deduplicationMinSize) {
actualSize += st.st_size;
unsharedSize += (st.st_nlink - 1) * st.st_size;
continue;
diff --git a/nix/libstore/local-store.hh b/nix/libstore/local-store.hh
index 9ba37219da..20d3c3c893 100644
--- a/nix/libstore/local-store.hh
+++ b/nix/libstore/local-store.hh
@@ -292,4 +292,7 @@ void canonicaliseTimestampAndPermissions(const Path & path);
MakeError(PathInUse, Error);
+/* Size below which a file is not considered for deduplication. */
+extern const size_t deduplicationMinSize;
+
}
diff --git a/nix/libstore/optimise-store.cc b/nix/libstore/optimise-store.cc
index eb303ab4c3..baca1a4890 100644
--- a/nix/libstore/optimise-store.cc
+++ b/nix/libstore/optimise-store.cc
@@ -15,6 +15,9 @@
namespace nix {
+/* Any file smaller than this is not considered for deduplication.
+ Keep in sync with (guix store deduplication). */
+const size_t deduplicationMinSize = 4096;
static void makeWritable(const Path & path)
{
@@ -105,12 +108,12 @@ void LocalStore::optimisePath_(OptimiseStats & stats, const Path & path, InodeHa
return;
}
- /* We can hard link regular files and maybe symlinks. */
- if (!S_ISREG(st.st_mode)
-#if CAN_LINK_SYMLINK
- && !S_ISLNK(st.st_mode)
-#endif
- ) return;
+ /* We can hard link regular files (and maybe symlinks), but do that only
+ for files larger than some threshold. This avoids adding too many
+ entries to '.links', which would slow down 'removeUnusedLinks' while
+ saving little space. */
+ if (!S_ISREG(st.st_mode) || ((size_t) st.st_size) < deduplicationMinSize)
+ return;
/* Sometimes SNAFUs can cause files in the store to be
modified, in particular when running programs as root under
diff --git a/tests/derivations.scm b/tests/derivations.scm
index cd165d1be6..4621098df3 100644
--- a/tests/derivations.scm
+++ b/tests/derivations.scm
@@ -1,5 +1,5 @@
;;; GNU Guix --- Functional package management for GNU
-;;; Copyright © 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020 Ludovic Courtès <ludo@gnu.org>
+;;; Copyright © 2012-2021 Ludovic Courtès <ludo@gnu.org>
;;;
;;; This file is part of GNU Guix.
;;;
@@ -170,11 +170,15 @@ (define prefix-len (string-length dir))
#f))))
(test-assert "identical files are deduplicated"
- (let* ((build1 (add-text-to-store %store "one.sh"
- "echo hello, world > \"$out\"\n"
+ ;; Note: DATA must be longer than %DEDUPLICATION-MINIMUM-SIZE.
+ (let* ((data (make-string 4500 #\a))
+ (build1 (add-text-to-store %store "one.sh"
+ (string-append "echo -n " data
+ " > \"$out\"\n")
'()))
(build2 (add-text-to-store %store "two.sh"
- "# Hey!\necho hello, world > \"$out\"\n"
+ (string-append "# Hey!\necho -n "
+ data " > \"$out\"\n")
'()))
(drv1 (derivation %store "foo"
%bash `(,build1)
@@ -187,7 +191,7 @@ (define prefix-len (string-length dir))
(file2 (derivation->output-path drv2)))
(and (valid-path? %store file1) (valid-path? %store file2)
(string=? (call-with-input-file file1 get-string-all)
- "hello, world\n")
+ data)
(= (stat:ino (lstat file1))
(stat:ino (lstat file2))))))))
diff --git a/tests/nar.scm b/tests/nar.scm
index ba4881caaa..bd2bf6e6e0 100644
--- a/tests/nar.scm
+++ b/tests/nar.scm
@@ -1,5 +1,5 @@
;;; GNU Guix --- Functional package management for GNU
-;;; Copyright © 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020 Ludovic Courtès <ludo@gnu.org>
+;;; Copyright © 2012-2021 Ludovic Courtès <ludo@gnu.org>
;;;
;;; This file is part of GNU Guix.
;;;
@@ -486,8 +486,9 @@ (define-values (port get-bytevector)
;; their mtime and permissions were not reset. Ensure that this bug is
;; gone.
(with-store store
- (let* ((text1 (random-text))
- (text2 (random-text))
+ ;; Note: TEXT1 and TEXT2 must be longer than %DEDUPLICATION-MINIMUM-SIZE.
+ (let* ((text1 (string-concatenate (make-list 100 (random-text))))
+ (text2 (string-concatenate (make-list 100 (random-text))))
(tree `("tree" directory
("a" regular (data ,text1))
("b" directory
diff --git a/tests/store-deduplication.scm b/tests/store-deduplication.scm
index b1c2d93bbd..b2b7c36622 100644
--- a/tests/store-deduplication.scm
+++ b/tests/store-deduplication.scm
@@ -1,5 +1,5 @@
;;; GNU Guix --- Functional package management for GNU
-;;; Copyright © 2018, 2020 Ludovic Courtès <ludo@gnu.org>
+;;; Copyright © 2018, 2020-2021 Ludovic Courtès <ludo@gnu.org>
;;;
;;; This file is part of GNU Guix.
;;;
@@ -30,13 +30,40 @@ (define-module (test-store-deduplication)
(test-begin "store-deduplication")
+(test-equal "deduplicate, below %deduplication-minimum-size"
+ (list #t (make-list 5 1))
+
+ (call-with-temporary-directory
+ (lambda (store)
+ ;; Note: DATA must be longer than %DEDUPLICATION-MINIMUM-SIZE.
+ (let ((data "Hello, world!")
+ (identical (map (lambda (n)
+ (string-append store "/" (number->string n)
+ "/a/b/c"))
+ (iota 5))))
+ (for-each (lambda (file)
+ (mkdir-p (dirname file))
+ (call-with-output-file file
+ (lambda (port)
+ (put-bytevector port (string->utf8 data)))))
+ identical)
+
+ (deduplicate store (nar-sha256 store) #:store store)
+
+ ;; (system (string-append "ls -lRia " store))
+ (list (= (length (delete-duplicates
+ (map (compose stat:ino stat) identical)))
+ (length identical))
+ (map (compose stat:nlink stat) identical))))))
+
(test-equal "deduplicate"
(cons* #t #f ;inode comparisons
2 (make-list 5 6)) ;'nlink' values
(call-with-temporary-directory
(lambda (store)
- (let ((data (string->utf8 "Hello, world!"))
+ ;; Note: DATA must be longer than %DEDUPLICATION-MINIMUM-SIZE.
+ (let ((data (string-concatenate (make-list 500 "Hello, world!")))
(identical (map (lambda (n)
(string-append store "/" (number->string n)
"/a/b/c"))
@@ -46,7 +73,7 @@ (define-module (test-store-deduplication)
(mkdir-p (dirname file))
(call-with-output-file file
(lambda (port)
- (put-bytevector port data))))
+ (put-bytevector port (string->utf8 data)))))
identical)
;; Make the parent of IDENTICAL read-only. This should not prevent
;; deduplication from inserting its hard link.
@@ -54,7 +81,7 @@ (define-module (test-store-deduplication)
(call-with-output-file unique
(lambda (port)
- (put-bytevector port (string->utf8 "This is unique."))))
+ (put-bytevector port (string->utf8 (string-reverse data)))))
(deduplicate store (nar-sha256 store) #:store store)
@@ -77,8 +104,10 @@ (define-module (test-store-deduplication)
(lambda (store)
(let ((true-link link)
(links 0)
- (data1 (string->utf8 "Hello, world!"))
- (data2 (string->utf8 "Hi, world!"))
+ (data1 (string->utf8
+ (string-concatenate (make-list 500 "Hello, world!"))))
+ (data2 (string->utf8
+ (string-concatenate (make-list 500 "Hi, world!"))))
(identical (map (lambda (n)
(string-append store "/" (number->string n)
"/a/b/c"))
diff --git a/tests/store.scm b/tests/store.scm
index 2150a0048c..5089909362 100644
--- a/tests/store.scm
+++ b/tests/store.scm
@@ -759,7 +759,9 @@ (define lst
(test-assert "substitute, deduplication"
(with-store s
- (let* ((c (random-text)) ; contents of the output
+ ;; Note: C must be longer than %DEDUPLICATION-MINIMUM-SIZE.
+ (let* ((c (string-concatenate
+ (make-list 100 (random-text)))) ; contents of the output
(g (package-derivation s %bootstrap-guile))
(d1 (build-expression->derivation s "substitute-me"
`(begin ,c (exit 1))
--
2.33.0
L
L
Ludovic Courtès wrote on 13 Nov 2021 22:45
Re: bug#24937: "deleting unused links" GC phase is too slow
(name . Maxim Cournoyer)(address . maxim.cournoyer@gmail.com)(address . 24937@debbugs.gnu.org)
87czn3buhg.fsf@gnu.org
Ludovic Courtès <ludo@gnu.org> skribis:

Toggle quote (6 lines)
> In a nutshell:
>
> • Files < 1KiB contribute to 0.3% of the space savings;
>
> • Files < 4KiB contribute to 2.5% of the space savings;

I get similar results on bayfront.guix.gnu.org (with 3.2M entries):
Attachment: bayfront-size.png
… and on guix.bordeaux.inria.fr (2.0M entries):
Attachment: guix-hpc4-size.png
Files < 4KiB represent between 60% and 75% of the /gnu/store/.links
entries here.

I’ve sent patches that implement a cutoff threshold at 4 KiB.

Thanks,
Ludo’.
L
L
Ludovic Courtès wrote on 16 Nov 2021 14:54
(address . 24937@debbugs.gnu.org)
87h7cc2ol6.fsf_-_@gnu.org
Hi,

Ludovic Courtès <ludo@gnu.org> skribis:

Toggle quote (10 lines)
> Files smaller than 4 KiB typically represent ~60% of the entries in
> /gnu/store/.links but only contribute to ~2.5% of the space savings
> afforded by deduplication.
>
> Not considering these files for deduplication speeds up file insertion
> in the store and, more importantly, leaves 'removeUnusedLinks' with
> fewer entries to traverse, thereby speeding it up proportionally.
>
> Partly fixes <https://issues.guix.gnu.org/24937>.

Pushed a variant of this as commit
472a0e82a52a3d5d841e1dfad6b13e26082a5750, with a threshold of 8 KiB.

Concretely, the number of .links entries shrinks by ~70%, from
2M to 700K on my laptop, and (presumably) from 64M to 19M on berlin.

I’ll deploy it within a few days on berlin. I hope the speedup will
reduce pressure there, though obviously it’ll still be an expensive
operation (but fundamentally I think it’ll always be linear in the size
of the store.)

I’m preparing an update of the ‘guix’ package to make this readily
available. When you deploy the new daemon, .links will be trimmed of
entries for files smaller than 8 KiB the first time you run ‘guix gc’.

Ludo’.
J
J
John Kehayias wrote on 22 Nov 2021 03:30
Re: "deleting unused links" GC phase is too slow
(name . 24937@debbugs.gnu.org)(address . 24937@debbugs.gnu.org)
QAT6Oklv6MJDMioc_c1DjSQKV_v-3AWmaSFX9mnO5dqeHMJ-v9W43_z4orvQ0cp5awD2GfvFlO8rBIh5jwnHyj1ZxXGMBUsTUsOf5nYq_l8=@protonmail.com
Hello,

A little late, but wanted to add my results here, from before the commit was made. I had reported some of this on IRC before and had some outlying results. Since then I finally did some generation deleting and gc-ing, though perhaps still have a bit of cruft. I've been on core-updates-frozen for a while, so keeping a lot from before making the branch switch, as well as lots of stuff piling up in trying to fix things or trying out changes in core-updates-frozen.

Anyway, attached are the plots from the above code. Running

ls -1A /gnu/store/.links | wc -l

showed 15776256 links at the time. Still quite a bit I think, but I've had 1.5-2x as much in the past, easily. (This had caused some earlier warnings on ext4 and enabling large_dir, which will make a system unbootable due to Grub not being up to speed on this old feature. I'm now on btrfs.)

John
Attachment: nlink.png
Attachment: nlink-small.png
Attachment: size-savings.png
Attachment: size.png
?