From 9da6fcbb3e6d720a4b575a48063ecf3240a44022 Mon Sep 17 00:00:00 2001
From: Eric Wong <e@80x24.org>
Date: Sat, 30 Apr 2016 02:57:40 +0000
Subject: daemon: graceful shutdown warning and limit removal

git clones may take longer than 30s, much longer...  So prepare
to wait almost indefinitely for sockets to timeout and document
the second signal behavior for immediate shutdown.

While we're at it, move parent death handling to a separate
class to avoid Danga::Socket->AddOtherFds, since that does not
allow proper handling the parent pipe being closed and would
actually misterminate a worker prematurely.  t/nntpd.t is update
to illustrate the failure with workers enabled.

We will work to keep memory usage low and let clients take their
time without interrupting them.
---
 lib/PublicInbox/Daemon.pm | 41 ++++++++++++++++++++++++++++++-----------
 1 file changed, 30 insertions(+), 11 deletions(-)

(limited to 'lib/PublicInbox/Daemon.pm')

diff --git a/lib/PublicInbox/Daemon.pm b/lib/PublicInbox/Daemon.pm
index c9594a37..8de7ff24 100644
--- a/lib/PublicInbox/Daemon.pm
+++ b/lib/PublicInbox/Daemon.pm
@@ -14,6 +14,7 @@ STDERR->autoflush(1);
 require Danga::Socket;
 require POSIX;
 require PublicInbox::Listener;
+require PublicInbox::ParentPipe;
 my @CMD;
 my $set_user;
 my (@cfg_listen, $stdout, $stderr, $group, $user, $pid_file, $daemonize);
@@ -161,16 +162,18 @@ sub daemonize () {
 	}
 }
 
-sub worker_quit () {
+
+sub worker_quit {
+	my ($reason) = @_;
 	# killing again terminates immediately:
 	exit unless @listeners;
 
 	$_->close foreach @listeners; # call Danga::Socket::close
 	@listeners = ();
+	$reason->close if ref($reason) eq 'PublicInbox::ParentPipe';
 
-	# give slow clients 30s to finish reading/writing whatever
-	Danga::Socket->AddTimer(30, sub { exit });
-
+	my $proc_name;
+	my $warn = 0;
 	# drop idle connections and try to quit gracefully
 	Danga::Socket->SetPostLoopCallback(sub {
 		my ($dmap, undef) = @_;
@@ -178,12 +181,23 @@ sub worker_quit () {
 
 		foreach my $s (values %$dmap) {
 			if ($s->can('busy') && $s->busy) {
-				$n = 1;
+				++$n;
 			} else {
 				# close as much as possible, early as possible
 				$s->close;
 			}
 		}
+		if ($n) {
+			if (($warn + 5) < time) {
+				warn "$$ quitting, $n client(s) left\n";
+				$warn = time;
+			}
+			unless (defined $proc_name) {
+				$proc_name = (split(/\s+/, $0))[0];
+				$proc_name =~ s!\A.*?([^/]+)\z!$1!;
+			}
+			$0 = "$proc_name quitting, $n client(s) left";
+		}
 		$n; # true: loop continues, false: loop breaks
 	});
 }
@@ -359,6 +373,7 @@ sub master_loop {
 	}
 	reopen_logs();
 	# main loop
+	my $quit = 0;
 	while (1) {
 		while (my $s = shift @caught) {
 			if ($s eq 'USR1') {
@@ -367,8 +382,8 @@ sub master_loop {
 			} elsif ($s eq 'USR2') {
 				upgrade();
 			} elsif ($s =~ /\A(?:QUIT|TERM|INT)\z/) {
-				# drops pipes and causes children to die
-				exit
+				exit if $quit++;
+				kill_workers($s);
 			} elsif ($s eq 'WINCH') {
 				$worker_processes = 0;
 			} elsif ($s eq 'HUP') {
@@ -390,6 +405,11 @@ sub master_loop {
 		}
 
 		my $n = scalar keys %pids;
+		if ($quit) {
+			exit if $n == 0;
+			$set_workers = $worker_processes = $n = 0;
+		}
+
 		if ($n > $worker_processes) {
 			while (my ($k, $v) = each %pids) {
 				kill('TERM', $k) if $v >= $worker_processes;
@@ -419,13 +439,12 @@ sub daemon_loop ($$) {
 	my $parent_pipe;
 	if ($worker_processes > 0) {
 		$refresh->(); # preload by default
-		$parent_pipe = master_loop(); # returns if in child process
-		my $fd = fileno($parent_pipe);
-		Danga::Socket->AddOtherFds($fd => *worker_quit);
+		my $fh = master_loop(); # returns if in child process
+		$parent_pipe = PublicInbox::ParentPipe->new($fh, *worker_quit);
 	} else {
 		reopen_logs();
 		$set_user->() if $set_user;
-		$SIG{USR2} = sub { worker_quit() if upgrade() };
+		$SIG{USR2} = sub { worker_quit('USR2') if upgrade() };
 		$refresh->();
 	}
 	$uid = $gid = undef;
-- 
cgit v1.2.3-24-ge0c7