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 # Internal interface for a single Xapian shard in V2 inboxes.
5 # See L<public-inbox-v2-format(5)> for more info on how we shard Xapian
6 package PublicInbox::SearchIdxShard;
9 use parent qw(PublicInbox::SearchIdx);
11 use IO::Handle (); # autoflush
13 use PublicInbox::Sigfd;
16 my ($class, $v2w, $shard) = @_; # v2w may be ExtSearchIdx
17 my $ibx = $v2w->{ibx};
18 my $self = $ibx ? $class->SUPER::new($ibx, 1, $shard)
19 : $class->eidx_shard_new($v2w, $shard);
20 # create the DB before forking:
22 $self->set_metadata_once;
24 $self->spawn_worker($v2w, $shard) if $v2w->{parallel};
29 my ($self, $v2w, $shard) = @_;
31 pipe($r, $w) or die "pipe failed: $!\n";
33 my $oldset = PublicInbox::Sigfd::block_signals();
35 defined $pid or die "fork failed: $!\n";
37 # these signals are localized in parent
38 $SIG{$_} = 'IGNORE' for (qw(TERM INT QUIT));
39 PublicInbox::Sigfd::sig_setmask($oldset);
40 my $bnote = $v2w->atfork_child;
41 close $w or die "failed to close: $!";
43 # F_SETPIPE_SZ = 1031 on Linux; increasing the pipe size here
44 # speeds V2Writable batch imports across 8 cores by nearly 20%
45 fcntl($r, 1031, 1048576) if $^O eq 'linux';
47 eval { shard_worker_loop($self, $v2w, $r, $shard, $bnote) };
48 die "worker $shard died: $@\n" if $@;
49 die "unexpected MM $self->{mm}" if $self->{mm};
52 PublicInbox::Sigfd::sig_setmask($oldset);
55 close $r or die "failed to close: $!";
61 my $n = read($r, my $bref, $len) or die "read: $!\n";
62 $n == $len or die "short read: $n != $len\n";
63 PublicInbox::Eml->new(\$bref);
66 # this reads all the writes to $self->{w} from the parent process
67 sub shard_worker_loop ($$$$$) {
68 my ($self, $v2w, $r, $shard, $bnote) = @_;
70 $self->begin_txn_lazy;
71 while (my $line = readline($r)) {
72 $v2w->{current_info} = "[$shard] $line";
73 if ($line eq "commit\n") {
74 $self->commit_txn_lazy;
75 } elsif ($line eq "close\n") {
77 } elsif ($line eq "barrier\n") {
78 $self->commit_txn_lazy;
79 # no need to lock < 512 bytes is atomic under POSIX
80 print $bnote "barrier $shard\n" or
81 die "write failed for barrier $!\n";
82 } elsif ($line =~ /\AD ([0-9]+)\n\z/s) {
83 $self->remove_by_docid($1 + 0);
84 } elsif ($line =~ s/\A\+X //) {
85 my ($len, $docid, $eidx_key) = split(/ /, $line, 3);
87 $self->add_eidx_info($docid, $eidx_key, eml($r, $len));
88 } elsif ($line =~ s/\A-X //) {
89 my ($len, $docid, $eidx_key) = split(/ /, $line, 3);
91 $self->remove_eidx_info($docid, $eidx_key,
93 } elsif ($line =~ s/\AO ([^\n]+)\n//) {
95 $over_fn =~ tr/\0/\n/;
96 $self->over_check(PublicInbox::Over->new($over_fn));
97 } elsif ($line =~ /\AE ([0-9]+)\n/) {
98 $self->reindex_docid($1 + 0);
102 if ($line =~ s/\AX=(.+)\0//) {
104 $v2w->{current_info} =~ s/\0/\\0 /;
106 # n.b. $mid may contain spaces(!)
107 my ($len, $bytes, $num, $oid, $ds, $ts, $tid, $mid)
108 = split(/ /, $line, 8);
109 $self->begin_txn_lazy;
118 }, 'PublicInbox::Smsg';
119 $smsg->{eidx_key} = $eidx_key if defined($eidx_key);
120 $self->add_message(eml($r, $len), $smsg);
127 my ($self, $msgref, $eml, $smsg, $ibx) = @_;
128 if (my $w = $self->{w}) {
129 my @ekey = $ibx ? ('X='.$ibx->eidx_key."\0") : ();
130 $msgref //= \($eml->as_string);
131 $smsg->{raw_bytes} //= length($$msgref);
132 # mid must be last, it can contain spaces (but not LF)
133 print $w @ekey, join(' ', @$smsg{qw(raw_bytes bytes
134 num blob ds ts tid mid)}),
135 "\n", $$msgref or die "failed to write shard $!\n";
138 undef($$msgref) if $msgref;
139 } else { # --xapian-only + --sequential-shard:
140 $eml = PublicInbox::Eml->new($msgref);
142 $self->begin_txn_lazy;
143 $smsg->{eidx_key} = $ibx->eidx_key if $ibx;
144 $self->add_message($eml, $smsg);
148 sub shard_add_eidx_info {
149 my ($self, $docid, $eidx_key, $eml) = @_;
150 if (my $w = $self->{w}) {
151 my $hdr = $eml->header_obj->as_string;
152 my $len = length($hdr);
153 print $w "+X $len $docid $eidx_key\n", $hdr or
154 die "failed to write shard: $!";
156 $self->add_eidx_info($docid, $eidx_key, $eml);
160 sub shard_remove_eidx_info {
161 my ($self, $docid, $eidx_key, $eml) = @_;
162 if (my $w = $self->{w}) {
163 my $hdr = $eml ? $eml->header_obj->as_string : '';
164 my $len = length($hdr);
165 print $w "-X $len $docid $eidx_key\n", $hdr or
166 die "failed to write shard: $!";
168 $self->remove_eidx_info($docid, $eidx_key, $eml);
173 close $_[0]->{w} or die "failed to close write pipe: $!\n";
178 if (my $w = $self->{w}) {
179 print $w "barrier\n" or die "failed to print: $!";
181 $self->commit_txn_lazy;
187 if (my $w = $self->{w}) {
188 print $w "commit\n" or die "failed to write commit: $!";
190 $self->commit_txn_lazy;
196 if (my $w = delete $self->{w}) {
197 my $pid = delete $self->{pid} or die "no process to wait on\n";
198 print $w "close\n" or die "failed to write to pid:$pid: $!\n";
199 close $w or die "failed to close pipe for pid:$pid: $!\n";
200 waitpid($pid, 0) == $pid or die "remote process did not finish";
201 $? == 0 or die ref($self)." pid:$pid exited with: $?";
203 die "transaction in progress $self\n" if $self->{txn};
204 $self->idx_release if $self->{xdb};
209 my ($self, $num) = @_;
210 if (my $w = $self->{w}) { # triggers remove_by_docid in a shard child
211 print $w "D $num\n" or die "failed to write remove $!";
212 } else { # same process
213 $self->remove_by_docid($num);
217 sub shard_over_check {
218 my ($self, $over) = @_;
219 if (my $w = $self->{w}) { # triggers remove_by_docid in a shard child
220 my ($over_fn) = $over->{dbh}->sqlite_db_filename;
221 $over_fn =~ tr/\n/\0/;
222 print $w "O $over_fn\n" or die "failed to write over $!";
224 $self->over_check($over);
228 sub shard_reindex_docid {
229 my ($self, $docid) = @_;
230 if (my $w = $self->{w}) {
231 print $w "E $docid\n" or die "failed to write to shard: $!";
233 $self->reindex_docid($docid);