1 # Copyright (C) 2018-2020 all contributors <meta@public-inbox.org>
2 # License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
4 # used to interface with a single Xapian shard in V2 repos.
5 # See L<public-inbox-v2-format(5)> for more info on how we shard Xapian
6 package PublicInbox::SearchIdxShard;
9 use base qw(PublicInbox::SearchIdx);
10 use IO::Handle (); # autoflush
13 my ($class, $v2writable, $shard) = @_;
14 my $self = $class->SUPER::new($v2writable->{-inbox}, 1, $shard);
15 # create the DB before forking:
18 $self->spawn_worker($v2writable, $shard) if $v2writable->{parallel};
23 my ($self, $v2w, $shard) = @_;
25 pipe($r, $w) or die "pipe failed: $!\n";
30 defined $pid or die "fork failed: $!\n";
32 my $bnote = $v2w->atfork_child;
33 close $w or die "failed to close: $!";
35 # F_SETPIPE_SZ = 1031 on Linux; increasing the pipe size here
36 # speeds V2Writable batch imports across 8 cores by nearly 20%
37 fcntl($r, 1031, 1048576) if $^O eq 'linux';
39 eval { shard_worker_loop($self, $v2w, $r, $shard, $bnote) };
40 die "worker $shard died: $@\n" if $@;
41 die "unexpected MM $self->{mm}" if $self->{mm};
46 close $r or die "failed to close: $!";
49 sub shard_worker_loop ($$$$$) {
50 my ($self, $v2w, $r, $shard, $bnote) = @_;
51 $0 = "pi-v2-shard[$shard]";
52 $self->begin_txn_lazy;
53 while (my $line = $r->getline) {
54 $v2w->{current_info} = "[$shard] $line";
55 if ($line eq "commit\n") {
56 $self->commit_txn_lazy;
57 } elsif ($line eq "close\n") {
59 } elsif ($line eq "barrier\n") {
60 $self->commit_txn_lazy;
61 # no need to lock < 512 bytes is atomic under POSIX
62 print $bnote "barrier $shard\n" or
63 die "write failed for barrier $!\n";
64 } elsif ($line =~ /\AD ([a-f0-9]{40,}) (.+)\n\z/s) {
65 my ($oid, $mid) = ($1, $2);
66 $self->begin_txn_lazy;
67 $self->remove_by_oid($oid, $mid);
70 my ($bytes, $num, $blob, $mid, $ds, $ts) =
72 $self->begin_txn_lazy;
73 my $n = read($r, my $msg, $bytes) or die "read: $!\n";
74 $n == $bytes or die "short read: $n != $bytes\n";
75 my $mime = PublicInbox::MIME->new(\$msg);
83 }, 'PublicInbox::Smsg';
84 $self->add_message($mime, $smsg);
90 # called by V2Writable
92 my ($self, $msgref, $mime, $smsg) = @_;
93 if (my $w = $self->{w}) {
94 print $w join(' ', @$smsg{qw(bytes num blob mid ds ts)}),
95 "\n", $$msgref or die "failed to write shard $!\n";
98 $self->begin_txn_lazy;
99 $self->add_message($mime, $smsg);
104 close $_[0]->{w} or die "failed to close write pipe: $!\n";
107 # called by V2Writable:
110 if (my $w = $self->{w}) {
111 print $w "barrier\n" or die "failed to print: $!";
113 $self->commit_txn_lazy;