-# Copyright (C) 2018-2019 all contributors <meta@public-inbox.org>
+# Copyright (C) 2018-2020 all contributors <meta@public-inbox.org>
# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
package PublicInbox::Xapcmd;
use strict;
use warnings;
-use PublicInbox::Spawn qw(which spawn);
+use PublicInbox::Spawn qw(which popen_rd);
use PublicInbox::Over;
-use PublicInbox::Search;
-use File::Temp qw(tempdir);
+use PublicInbox::SearchIdx;
+use File::Temp 0.19 (); # ->newdir
use File::Path qw(remove_tree);
use File::Basename qw(dirname);
+use POSIX ();
# support testing with dev versions of Xapian which installs
# commands with a version number suffix (e.g. "xapian-compact-1.5")
our $XAPIAN_COMPACT = $ENV{XAPIAN_COMPACT} || 'xapian-compact';
our @COMPACT_OPT = qw(jobs|j=i quiet|q blocksize|b=s no-full|n fuller|F);
-sub commit_changes ($$$) {
- my ($ibx, $tmp, $opt) = @_;
- my $new_parts = $opt->{reshard};
+sub commit_changes ($$$$) {
+ my ($ibx, $im, $tmp, $opt) = @_;
+ my $reshard = $opt->{reshard};
my $reindex = $opt->{reindex};
- my $im = $ibx->importer(0);
- $im->lock_acquire if !$opt->{-coarse_lock};
$SIG{INT} or die 'BUG: $SIG{INT} not handled';
- my @old_part;
+ my @old_shard;
- while (my ($old, $new) = each %$tmp) {
+ while (my ($old, $newdir) = each %$tmp) {
+ next if $old eq ''; # no invalid paths
my @st = stat($old);
if (!@st && !defined($opt->{reshard})) {
die "failed to stat($old): $!";
}
+ my $new = $newdir->dirname if defined($newdir);
my $over = "$old/over.sqlite3";
if (-f $over) { # only for v1, v2 over is untouched
defined $new or die "BUG: $over exists when culling v2";
$over = undef;
}
- if (!defined($new)) { # culled partition
- push @old_part, $old;
+ if (!defined($new)) { # culled shard
+ push @old_shard, $old;
next;
}
die "failed to remove $prev: $!\n";
}
}
- remove_tree(@old_part);
- $tmp->done;
+ remove_tree(@old_shard);
+ $tmp = undef;
if (!$opt->{-coarse_lock}) {
$opt->{-skip_lock} = 1;
- if ($im->can('count_partitions')) {
+ if ($im->can('count_shards')) {
my $pr = $opt->{-progress};
- my $n = $im->count_partitions;
- if (defined $new_parts && $n != $new_parts) {
+ my $n = $im->count_shards;
+ if (defined $reshard && $n != $reshard) {
die
-"BUG: counted $n partitions after repartioning to $new_parts";
+"BUG: counted $n shards after resharding to $reshard";
}
- my $prev = $im->{partitions};
+ my $prev = $im->{shards};
if ($pr && $prev != $n) {
- $pr->("partition count changed: $prev => $n\n");
- $im->{partitions} = $n;
+ $pr->("shard count changed: $prev => $n\n");
+ $im->{shards} = $n;
}
}
- PublicInbox::Admin::index_inbox($ibx, $opt);
- # implicit lock_release
- } else {
- $im->lock_release;
+ PublicInbox::Admin::index_inbox($ibx, $im, $opt);
}
}
defined(my $pid = fork) or die "fork: $!";
return $pid if $pid > 0;
$cb->($args, $opt);
- exit 0;
+ POSIX::_exit(0);
}
sub runnable_or_die ($) {
which($exe) or die "$exe not found in PATH\n";
}
-sub prepare_reindex ($$) {
- my ($ibx, $reindex) = @_;
- if ($ibx->{version} == 1) {
+sub prepare_reindex ($$$) {
+ my ($ibx, $im, $reindex) = @_;
+ if ($ibx->version == 1) {
my $dir = $ibx->search->xdir(1);
- my $xdb = Search::Xapian::Database->new($dir);
+ my $xdb = $PublicInbox::Search::X{Database}->new($dir);
if (my $lc = $xdb->get_metadata('last_commit')) {
$reindex->{from} = $lc;
}
} else { # v2
- my $v2w = $ibx->importer(0);
my $max;
- $v2w->git_dir_latest(\$max) or return;
+ $im->git_dir_latest(\$max) or return;
my $from = $reindex->{from};
my $mm = $ibx->mm;
my $v = PublicInbox::Search::SCHEMA_VERSION();
}
sub process_queue {
- my ($queue, $cb, $max, $opt) = @_;
+ my ($queue, $cb, $opt) = @_;
+ my $max = $opt->{jobs} || scalar(@$queue);
if ($max <= 1) {
while (defined(my $args = shift @$queue)) {
$cb->($args, $opt);
}
}
-sub run {
- my ($ibx, $task, $opt) = @_; # task = 'cpdb' or 'compact'
- my $cb = \&${\"PublicInbox::Xapcmd::$task"};
- PublicInbox::Admin::progress_prepare($opt ||= {});
- my $dir = $ibx->{mainrepo} or die "no mainrepo in inbox\n";
- runnable_or_die($XAPIAN_COMPACT) if $opt->{compact};
- my $reindex; # v1:{ from => $x40 }, v2:{ from => [ $x40, $x40, .. ] } }
- my $from; # per-epoch ranges
+sub setup_signals () {
+ # http://www.tldp.org/LDP/abs/html/exitcodes.html
+ $SIG{INT} = sub { exit(130) };
+ $SIG{HUP} = $SIG{PIPE} = $SIG{TERM} = sub { exit(1) };
+}
- if (!$opt->{-coarse_lock}) {
- $reindex = $opt->{reindex} = {};
- $from = $reindex->{from} = [];
- require Search::Xapian::WritableDatabase;
- }
+sub prepare_run {
+ my ($ibx, $opt) = @_;
+ my $tmp = {}; # old shard dir => File::Temp->newdir object or undef
+ my @queue; # ([old//src,newdir]) - list of args for cpdb() or compact()
- $ibx->umask_prepare;
my $old = $ibx->search->xdir(1);
-d $old or die "$old does not exist\n";
-
- my $tmp = PublicInbox::Xtmpdirs->new;
- my $v = $ibx->{version} ||= 1;
- my @q;
- my $new_parts = $opt->{reshard};
- if (defined $new_parts && $new_parts <= 0) {
+ my $reshard = $opt->{reshard};
+ if (defined $reshard && $reshard <= 0) {
die "--reshard must be a positive number\n";
}
# we want temporary directories to be as deep as possible,
- # so v2 partitions can keep "xap$SCHEMA_VERSION" on a separate FS.
- if ($v == 1) {
- if (defined $new_parts) {
+ # so v2 shards can keep "xap$SCHEMA_VERSION" on a separate FS.
+ if ($ibx->version == 1) {
+ if (defined $reshard) {
warn
-"--reshard=$new_parts ignored for v1 $ibx->{mainrepo}\n";
+"--reshard=$reshard ignored for v1 $ibx->{inboxdir}\n";
}
- my $old_parent = dirname($old);
- same_fs_or_die($old_parent, $old);
+ my $dir = dirname($old);
+ same_fs_or_die($dir, $old);
my $v = PublicInbox::Search::SCHEMA_VERSION();
- my $wip = tempdir("xapian$v-XXXXXXXX", DIR => $old_parent);
+ my $wip = File::Temp->newdir("xapian$v-XXXXXXXX", DIR => $dir);
$tmp->{$old} = $wip;
- push @q, [ $old, $wip ];
+ push @queue, [ $old, $wip ];
} else {
opendir my $dh, $old or die "Failed to opendir $old: $!\n";
- my @old_parts;
+ my @old_shards;
while (defined(my $dn = readdir($dh))) {
if ($dn =~ /\A[0-9]+\z/) {
- push @old_parts, $dn;
+ push @old_shards, $dn;
} elsif ($dn eq '.' || $dn eq '..') {
} elsif ($dn =~ /\Aover\.sqlite3/) {
} else {
warn "W: skipping unknown dir: $old/$dn\n"
}
}
- die "No Xapian parts found in $old\n" unless @old_parts;
+ die "No Xapian shards found in $old\n" unless @old_shards;
- my ($src, $max_part);
- if (!defined($new_parts) || $new_parts == scalar(@old_parts)) {
+ my ($src, $max_shard);
+ if (!defined($reshard) || $reshard == scalar(@old_shards)) {
# 1:1 copy
- $max_part = scalar(@old_parts) - 1;
+ $max_shard = scalar(@old_shards) - 1;
} else {
# M:N copy
- $max_part = $new_parts - 1;
- $src = [ map { "$old/$_" } @old_parts ];
+ $max_shard = $reshard - 1;
+ $src = [ map { "$old/$_" } @old_shards ];
}
- foreach my $dn (0..$max_part) {
+ foreach my $dn (0..$max_shard) {
my $tmpl = "$dn-XXXXXXXX";
- my $wip = tempdir($tmpl, DIR => $old);
- same_fs_or_die($old, $wip);
+ my $wip = File::Temp->newdir($tmpl, DIR => $old);
+ same_fs_or_die($old, $wip->dirname);
my $cur = "$old/$dn";
- push @q, [ $src // $cur , $wip ];
+ push @queue, [ $src // $cur , $wip ];
$tmp->{$cur} = $wip;
}
- # mark old parts to be unlinked
+ # mark old shards to be unlinked
if ($src) {
$tmp->{$_} ||= undef for @$src;
}
}
- my $im = $ibx->importer(0);
- my $max = $opt->{jobs} || scalar(@q);
+ ($tmp, \@queue);
+}
+
+sub check_compact () { runnable_or_die($XAPIAN_COMPACT) }
+
+sub run {
+ my ($ibx, $task, $opt) = @_; # task = 'cpdb' or 'compact'
+ my $cb = \&${\"PublicInbox::Xapcmd::$task"};
+ PublicInbox::Admin::progress_prepare($opt ||= {});
+ defined(my $dir = $ibx->{inboxdir}) or die "no inboxdir defined\n";
+ -d $dir or die "inboxdir=$dir does not exist\n";
+ check_compact() if $opt->{compact};
+ my $reindex; # v1:{ from => $x40 }, v2:{ from => [ $x40, $x40, .. ] } }
+
+ if (!$opt->{-coarse_lock}) {
+ $reindex = $opt->{reindex} = {};
+ $reindex->{from} = []; # per-epoch ranges
+ require PublicInbox::SearchIdx;
+ PublicInbox::SearchIdx::load_xapian_writable();
+ }
+
+ local %SIG = %SIG;
+ setup_signals();
+ $ibx->umask_prepare;
$ibx->with_umask(sub {
+ my $im = $ibx->importer(0);
$im->lock_acquire;
+ my ($tmp, $queue) = prepare_run($ibx, $opt);
# fine-grained locking if we prepare for reindex
if (!$opt->{-coarse_lock}) {
- prepare_reindex($ibx, $reindex);
+ prepare_reindex($ibx, $im, $reindex);
$im->lock_release;
}
- delete($ibx->{$_}) for (qw(mm over search)); # cleanup
- process_queue(\@q, $cb, $max, $opt);
- commit_changes($ibx, $tmp, $opt);
+ $ibx->cleanup;
+ process_queue($queue, $cb, $opt);
+ $im->lock_acquire if !$opt->{-coarse_lock};
+ commit_changes($ibx, $im, $tmp, $opt);
});
}
sub cpdb_retryable ($$) {
my ($src, $pfx) = @_;
- if (ref($@) eq 'Search::Xapian::DatabaseModifiedError') {
+ if (ref($@) =~ /\bDatabaseModifiedError\b/) {
warn "$pfx Xapian DB modified, reopening and retrying\n";
$src->reopen;
return 1;
# xapian-compact wrapper
sub compact ($$) {
my ($args, $opt) = @_;
- my ($src, $dst) = @$args;
- my ($r, $w);
+ my ($src, $newdir) = @$args;
+ my $dst = ref($newdir) ? $newdir->dirname : $newdir;
my $pfx = $opt->{-progress_pfx} ||= progress_pfx($src);
my $pr = $opt->{-progress};
my $rdr = {};
defined(my $dfd = $opt->{$fd}) or next;
$rdr->{$fd} = $dfd;
}
- $rdr->{1} = fileno($w) if $pr && pipe($r, $w);
# we rely on --no-renumber to keep docids synched to NNTP
my $cmd = [ $XAPIAN_COMPACT, '--no-renumber' ];
}
$pr->("$pfx `".join(' ', @$cmd)."'\n") if $pr;
push @$cmd, $src, $dst;
- my $pid = spawn($cmd, undef, $rdr);
- if ($pr) {
- close $w or die "close: \$w: $!";
- foreach (<$r>) {
+ my $rd = popen_rd($cmd, undef, $rdr);
+ while (<$rd>) {
+ if ($pr) {
s/\r/\r$pfx /g;
$pr->("$pfx $_");
}
}
- my $rp = waitpid($pid, 0);
- if ($? || $rp != $pid) {
- die join(' ', @$cmd)." failed: $? (pid=$pid, reaped=$rp)\n";
- }
+ close $rd or die join(' ', @$cmd)." failed: $?n";
}
sub cpdb_loop ($$$;$$) {
- my ($src, $dst, $pr_data, $cur_part, $new_parts) = @_;
+ my ($src, $dst, $pr_data, $cur_shard, $reshard) = @_;
my ($pr, $fmt, $nr, $pfx);
if ($pr_data) {
$pr = $pr_data->{pr};
eval {
for (; $it != $end; $it++) {
my $docid = $it->get_docid;
- if (defined $new_parts) {
- my $dst_part = $docid % $new_parts;
- next if $dst_part != $cur_part;
+ if (defined $reshard) {
+ my $dst_shard = $docid % $reshard;
+ next if $dst_shard != $cur_shard;
}
my $doc = $src->get_document($docid);
$dst->replace_document($docid, $doc);
# to the overhead of Perl.
sub cpdb ($$) {
my ($args, $opt) = @_;
- my ($old, $new) = @$args;
- my ($src, $cur_part);
- my $new_parts;
+ my ($old, $newdir) = @$args;
+ my $new = $newdir->dirname;
+ my ($src, $cur_shard);
+ my $reshard;
+ PublicInbox::SearchIdx::load_xapian_writable() or die;
+ my $XapianDatabase = $PublicInbox::Search::X{Database};
if (ref($old) eq 'ARRAY') {
- ($cur_part) = ($new =~ m!xap[0-9]+/([0-9]+)\b!);
- defined $cur_part or
- die "BUG: could not extract partition # from $new";
- $new_parts = $opt->{reshard};
- defined $new_parts or die 'BUG: got array src w/o --partition';
+ ($cur_shard) = ($new =~ m!xap[0-9]+/([0-9]+)\b!);
+ defined $cur_shard or
+ die "BUG: could not extract shard # from $new";
+ $reshard = $opt->{reshard};
+ defined $reshard or die 'BUG: got array src w/o --reshard';
- # repartitioning, M:N copy means have full read access
+ # resharding, M:N copy means have full read access
foreach (@$old) {
if ($src) {
- my $sub = Search::Xapian::Database->new($_);
+ my $sub = $XapianDatabase->new($_);
$src->add_database($sub);
} else {
- $src = Search::Xapian::Database->new($_);
+ $src = $XapianDatabase->new($_);
}
}
} else {
- $src = Search::Xapian::Database->new($old);
+ $src = $XapianDatabase->new($old);
}
- my ($xtmp, $tmp);
+ my ($tmp, $ft);
+ local %SIG = %SIG;
if ($opt->{compact}) {
- my $newdir = dirname($new);
- same_fs_or_die($newdir, $new);
- $tmp = tempdir("$new.compact-XXXXXX", DIR => $newdir);
- $xtmp = PublicInbox::Xtmpdirs->new;
- $xtmp->{$new} = $tmp;
+ my $dir = dirname($new);
+ same_fs_or_die($dir, $new);
+ $ft = File::Temp->newdir("$new.compact-XXXXXX", DIR => $dir);
+ setup_signals();
+ $tmp = $ft->dirname;
} else {
$tmp = $new;
}
# like copydatabase(1), be sure we don't overwrite anything in case
# of other bugs:
- my $creat = Search::Xapian::DB_CREATE();
- my $dst = Search::Xapian::WritableDatabase->new($tmp, $creat);
+ my $creat = eval($PublicInbox::Search::Xap.'::DB_CREATE()');
+ die if $@;
+ my $XapianWritableDatabase = $PublicInbox::Search::X{WritableDatabase};
+ my $dst = $XapianWritableDatabase->new($tmp, $creat);
my $pr = $opt->{-progress};
my $pfx = $opt->{-progress_pfx} = progress_pfx($new);
my $pr_data = { pr => $pr, pfx => $pfx, nr => 0 } if $pr;
my $lc = $src->get_metadata('last_commit');
$dst->set_metadata('last_commit', $lc) if $lc;
- # only the first xapian partition (0) gets 'indexlevel'
+ # only the first xapian shard (0) gets 'indexlevel'
if ($new =~ m!(?:xapian[0-9]+|xap[0-9]+/0)\b!) {
my $l = $src->get_metadata('indexlevel');
if ($l eq 'medium') {
if ($pr_data) {
my $tot = $src->get_doccount;
- # we can only estimate when repartitioning,
+ # we can only estimate when resharding,
# because removed spam causes slight imbalance
my $est = '';
- if (defined $cur_part && $new_parts > 1) {
- $tot = int($tot/$new_parts);
+ if (defined $cur_shard && $reshard > 1) {
+ $tot = int($tot/$reshard);
$est = 'around ';
}
my $fmt = "$pfx % ".length($tot)."u/$tot\n";
};
} while (cpdb_retryable($src, $pfx));
- if (defined $new_parts) {
+ if (defined $reshard) {
# we rely on document IDs matching NNTP article number,
- # so we can't have the combined DB support rewriting
+ # so we can't have the Xapian sharding DB support rewriting
# document IDs. Thus we iterate through each shard
# individually.
$src = undef;
foreach (@$old) {
- my $old = Search::Xapian::Database->new($_);
- cpdb_loop($old, $dst, $pr_data, $cur_part, $new_parts);
+ my $old = $XapianDatabase->new($_);
+ cpdb_loop($old, $dst, $pr_data, $cur_shard, $reshard);
}
} else {
cpdb_loop($src, $dst, $pr_data);
}
$pr->(sprintf($pr_data->{fmt}, $pr_data->{nr})) if $pr;
- return unless $xtmp;
+ return unless $opt->{compact};
$src = $dst = undef; # flushes and closes
# since $dst isn't readable by HTTP or NNTP clients, yet:
compact([ $tmp, $new ], $opt);
remove_tree($tmp) or die "failed to remove $tmp: $!\n";
- $xtmp->done;
-}
-
-# slightly easier-to-manage manage than END{} blocks
-package PublicInbox::Xtmpdirs;
-use strict;
-use warnings;
-use File::Path qw(remove_tree);
-my %owner;
-
-sub new {
- # http://www.tldp.org/LDP/abs/html/exitcodes.html
- $SIG{INT} = sub { exit(130) };
- $SIG{HUP} = $SIG{PIPE} = $SIG{TERM} = sub { exit(1) };
- my $self = bless {}, $_[0]; # old partition => new (tmp) partition
- $owner{"$self"} = $$;
- $self;
-}
-
-sub done {
- my ($self) = @_;
- delete $owner{"$self"};
-
- my %known_pids;
- $known_pids{$_}++ foreach values %owner;
- if (!$known_pids{$$}) {
- $SIG{INT} = $SIG{HUP} = $SIG{PIPE} = $SIG{TERM} = 'DEFAULT';
- }
- %$self = ();
-}
-
-sub DESTROY {
- my ($self) = @_;
- my $owner_pid = delete $owner{"$self"} or return;
- return if $owner_pid != $$;
- foreach my $new (values %$self) {
- defined $new or next; # may be undef if repartitioning
- remove_tree($new) unless -d "$new/old";
- }
- done($self);
}
1;