From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on dcvr.yhbt.net X-Spam-Level: X-Spam-ASN: X-Spam-Status: No, score=-4.0 required=3.0 tests=ALL_TRUSTED,BAYES_00 shortcircuit=no autolearn=ham autolearn_force=no version=3.4.2 Received: from localhost (dcvr.yhbt.net [127.0.0.1]) by dcvr.yhbt.net (Postfix) with ESMTP id F39511F466 for ; Sun, 26 Jan 2020 23:41:02 +0000 (UTC) From: Eric Wong To: meta@public-inbox.org Subject: [PATCH] v2writable: newest epochs go first in alternates Date: Sun, 26 Jan 2020 23:41:02 +0000 Message-Id: <20200126234102.22958-1-e@yhbt.net> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit List-Id: New epochs are the most likely to have loose objects. git won't be able to take advantage of pack indices and needs to scan every alternate for the loose object via open/openat syscalls. Those syscalls will add up some day when we've got hundreds or thousands of epochs. --- lib/PublicInbox/V2Writable.pm | 42 ++++++++++++++++++++++++++--------- t/v2writable.t | 24 ++++++++++++++++++++ 2 files changed, 55 insertions(+), 11 deletions(-) diff --git a/lib/PublicInbox/V2Writable.pm b/lib/PublicInbox/V2Writable.pm index 9073d9ef..8ad71b54 100644 --- a/lib/PublicInbox/V2Writable.pm +++ b/lib/PublicInbox/V2Writable.pm @@ -663,23 +663,43 @@ sub fill_alternates ($$) { unless (-d $all) { PublicInbox::Import::init_bare($all); } - my $alt = "$all/objects/info/alternates"; - my %alts; - my @add; + my $info_dir = "$all/objects/info"; + my $alt = "$info_dir/alternates"; + my (%alt, $new); + my $mode = 0644; if (-e $alt) { open(my $fh, '<', $alt) or die "open < $alt: $!\n"; - %alts = map { chomp; $_ => 1 } (<$fh>); + $mode = (stat($fh))[2] & 07777; + + # we assign a sort score to every alternate and favor + # the newest (highest numbered) one when we + my $score; + my $other = 0; # in case admin adds non-epoch repos + %alt = map {; + if (m!\A\Q../../\E([0-9]+)\.git/objects\z!) { + $score = $1 + 0; + } else { + $score = --$other; + } + $_ => $score; + } split(/\n+/, do { local $/; <$fh> }); } + foreach my $i (0..$epoch) { my $dir = "../../git/$i.git/objects"; - push @add, $dir if !$alts{$dir} && -d "$pfx/$i.git"; - } - return unless @add; - open my $fh, '>>', $alt or die "open >> $alt: $!\n"; - foreach my $dir (@add) { - print $fh "$dir\n" or die "print >> $alt: $!\n"; + if (!exists($alt{$dir}) && -d "$pfx/$i.git") { + $alt{$dir} = $i; + $new = 1; + } } - close $fh or die "close $alt: $!\n"; + return unless $new; + + my ($fh, $tmp) = tempfile('alt-XXXXXXXX', DIR => $info_dir); + print $fh join("\n", sort { $alt{$b} <=> $alt{$a} } keys %alt), "\n" + or die "print $tmp: $!\n"; + chmod($mode, $fh) or die "fchmod $tmp: $!\n"; + close $fh or die "close $tmp $!\n"; + rename($tmp, $alt) or die "rename $tmp => $alt: $!\n"; } sub git_init { diff --git a/t/v2writable.t b/t/v2writable.t index 63f27f96..7ceaa6e4 100644 --- a/t/v2writable.t +++ b/t/v2writable.t @@ -284,4 +284,28 @@ eval { }; ok($@, 'V2Writable fails on non-existent dir'); +{ + my $v2w = PublicInbox::V2Writable->new($tmp, 1); + ok($v2w, 'creat flag works'); + $v2w->{parallel} = 0; + $v2w->init_inbox(0); + my $alt = "$tmp->{inboxdir}/all.git/objects/info/alternates"; + open my $fh, '>>', $alt or die $!; + print $fh "$inboxdir/all.git/objects\n" or die $!; + chmod(0664, $fh) or die "fchmod: $!"; + close $fh or die $!; + open $fh, '<', $alt or die $!; + my $before = do { local $/; <$fh> }; + + ok($v2w->git_init(3), 'init a new epoch'); + open $fh, '<', $alt or die $!; + my $after = do { local $/; <$fh> }; + ok(index($after, $before) > 0, + 'old contents preserved after adding epoch'); + like($after, qr!\A[^\n]+?/3\.git/objects\n!s, + 'first line is newest epoch'); + my $mode = (stat($alt))[2] & 07777; + is($mode, 0664, sprintf('0%03o', $mode).' is 0664'); +} + done_testing();