loncom/loncnew - diff

Return to loncnew CVS log

Up to [LON-CAPA] / loncom

Diff for /loncom/loncnew between versions 1.6 and 1.23

version 1.6, 2003/05/05 23:40:27	version 1.23, 2003/09/15 09:24:49
Line 7	Line 7
# Copyright Michigan State University Board of Trustees	# Copyright Michigan State University Board of Trustees
#	#
# This file is part of the LearningOnline Network with CAPA (LON-CAPA).	# This file is part of the LearningOnline Network with CAPA (LON-CAPA).
#	## LON-CAPA is free software; you can redistribute it and/or modify
# LON-CAPA is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by	# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or	# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.	# (at your option) any later version.
Line 27	Line 26
# http://www.lon-capa.org/	# http://www.lon-capa.org/
#	#
#	#
# new lonc handles n requestors spread out bver m connections to londs.	# new lonc handles n request out bver m connections to londs.
# This module is based on the Event class.	# This module is based on the Event class.
# Development iterations:	# Development iterations:
# - Setup basic event loop. (done)	# - Setup basic event loop. (done)
# - Add timer dispatch. (done)	# - Add timer dispatch. (done)
# - Add ability to accept lonc UNIX domain sockets. (done)	# - Add ability to accept lonc UNIX domain sockets. (done)
# - Add ability to create/negotiate lond connections (done).	# - Add ability to create/negotiate lond connections (done).
# - Add general logic for dispatching requests and timeouts.	# - Add general logic for dispatching requests and timeouts. (done).
# - Add support for the lonc/lond requests.	# - Add support for the lonc/lond requests. (done).
# - Add logging/status monitoring.	# - Add logging/status monitoring.
# - Add Signal handling - HUP restarts. USR1 status report.	# - Add Signal handling - HUP restarts. USR1 status report.
# - Add Configuration file I/O	# - Add Configuration file I/O (done).
# - Add Pending request processing on startup.
# - Add management/status request interface.	# - Add management/status request interface.
	# - Add deferred request capability. (done)
	# - Detect transmission timeouts.
	#

	# Change log:
	# $Log$
	# Revision 1.23 2003/09/15 09:24:49 foxr
	# Add use strict and fix all the fallout from that.
	#
	# Revision 1.22 2003/09/02 10:34:47 foxr
	# - Fix errors in host dead detection logic (too many cases where the
	# retries left were not getting incremented or just not checked).
	# - Added some additional status to the ps axuww display:
	# o Remaining retries on a host.
	# o >>> DEAD <<< indicator if I've given up on a host.
	# - Tested the SIGHUP will reset the retries remaining count (thanks to
	# the above status stuff, and get allow the loncnew to re-try again
	# on the host (thanks to the log).
	#
	# Revision 1.21 2003/08/26 09:19:51 foxr
	# How embarrassing... put in the SocketTimeout function in loncnew and forgot
	# to actually hook it into the LondTransaction. Added this to MakeLondConnection
	# where it belongs... hopefully transactions (not just connection attempts) will
	# timeout more speedily than the socket errors will catch it.
	#
	# Revision 1.20 2003/08/25 18:48:11 albertel
	# - fixing a forgotten ;
	#
	# Revision 1.19 2003/08/19 09:31:46 foxr
	# Get socket directory from configuration rather than the old hard coded test
	# way that I forgot to un-hard code.
	#
	# Revision 1.18 2003/08/06 09:52:29 foxr
	# Also needed to remember to fail in-flight transactions if their sends fail.
	#
	# Revision 1.17 2003/08/03 00:44:31 foxr
	# 1. Correct handling of connection failure: Assume it means the host is
	# unreachable and fail all of the queued transactions. Note that the
	# inflight transactions should fail on their own time due either to timeout
	# or send/receive failures.
	# 2. Correct handling of logs for forced death signals. Pull the signal
	# from the event watcher.
	#
	# Revision 1.16 2003/07/29 02:33:05 foxr
	# Add SIGINT processing to child processes to toggle annoying trace mode
	# on/off.. will try to use this to isolate the compute boud process issue.
	#
	# Revision 1.15 2003/07/15 02:07:05 foxr
	# Added code for lonc/lond transaction timeouts. Who knows if it works right.
	# The intent is for a timeout to fail any transaction in progress and kill
	# off the sockt that timed out.
	#
	# Revision 1.14 2003/07/03 02:10:18 foxr
	# Get all of the signals to work correctly.
	#
	# Revision 1.13 2003/07/02 01:31:55 foxr
	# Added kill -HUP logic (restart).
	#
	# Revision 1.11 2003/06/25 01:54:44 foxr
	# Fix more problems with transaction failure.
	#
	# Revision 1.10 2003/06/24 02:46:04 foxr
	# Put a limit on the number of times we'll retry a connection.
	# Start getting the signal stuff put in as well...note that need to get signals
	# going or else the client will permanently give up on dead servers.
	#
	# Revision 1.9 2003/06/13 02:38:43 foxr
	# Add logging in 'expected format'
	#
	# Revision 1.8 2003/06/11 02:04:35 foxr
	# Support delayed transactions... this is done uniformly by encapsulating
	# transactions in an object ... a LondTransaction that is implemented by
	# LondTransaction.pm
	#
	# Revision 1.7 2003/06/03 01:59:39 foxr
	# complete coding to support deferred transactions.
	#
	#
	use strict;
use lib "/home/httpd/lib/perl/";	use lib "/home/httpd/lib/perl/";
use lib "/home/foxr/newloncapa/types";	use lib "/home/foxr/newloncapa/types";
use Event qw(:DEFAULT );	use Event qw(:DEFAULT );
use POSIX qw(:signal_h);	use POSIX qw(:signal_h);
	use POSIX;
use IO::Socket;	use IO::Socket;
use IO::Socket::INET;	use IO::Socket::INET;
use IO::Socket::UNIX;	use IO::Socket::UNIX;
	use IO::File;
use IO::Handle;	use IO::Handle;
use Socket;	use Socket;
use Crypt::IDEA;	use Crypt::IDEA;
use LONCAPA::Queue;	use LONCAPA::Queue;
use LONCAPA::Stack;	use LONCAPA::Stack;
use LONCAPA::LondConnection;	use LONCAPA::LondConnection;
	use LONCAPA::LondTransaction;
use LONCAPA::Configuration;	use LONCAPA::Configuration;
use LONCAPA::HashIterator;	use LONCAPA::HashIterator;

print "Loncnew starting\n";

#	#
# Disable all signals we might receive from outside for now.	# Disable all signals we might receive from outside for now.
#	#
$SIG{QUIT} = IGNORE;	#$SIG{QUIT} = IGNORE;
$SIG{HUP} = IGNORE;	#$SIG{HUP} = IGNORE;
$SIG{USR1} = IGNORE;	#$SIG{USR1} = IGNORE;
$SIG{INT} = IGNORE;	#$SIG{INT} = IGNORE;
$SIG{CHLD} = IGNORE;	#$SIG{CHLD} = IGNORE;
$SIG{__DIE__} = IGNORE;	#$SIG{__DIE__} = IGNORE;


# Read the httpd configuration file to get perl variables	# Read the httpd configuration file to get perl variables
Line 83 my %perlvar = %{$perlvarref};	Line 161 my %perlvar = %{$perlvarref};
my %ChildHash; # by pid -> host.	my %ChildHash; # by pid -> host.


my $MaxConnectionCount = 5; # Will get from config later.	my $MaxConnectionCount = 10; # Will get from config later.
my $ClientConnection = 0; # Uniquifier for client events.	my $ClientConnection = 0; # Uniquifier for client events.

my $DebugLevel = 5;	my $DebugLevel = 0;
	my $NextDebugLevel= 10; # So Sigint can toggle this.
my $IdleTimeout= 3600; # Wait an hour before pruning connections.	my $IdleTimeout= 3600; # Wait an hour before pruning connections.

#	#
# The variables below are only used by the child processes.	# The variables below are only used by the child processes.
#	#
my $RemoteHost; # Name of host child is talking to.	my $RemoteHost; # Name of host child is talking to.
my $UnixSocketDir= "/home/httpd/sockets";	my $UnixSocketDir= $perlvar{'lonSockDir'};
my $IdleConnections = Stack->new(); # Set of idle connections	my $IdleConnections = Stack->new(); # Set of idle connections
my %ActiveConnections; # Connections to the remote lond.	my %ActiveConnections; # Connections to the remote lond.
my %ActiveTransactions; # Transactions in flight.	my %ActiveTransactions; # LondTransactions in flight.
my %ActiveClients; # Serial numbers of active clients by socket.	my %ActiveClients; # Serial numbers of active clients by socket.
my $WorkQueue = Queue->new(); # Queue of pending transactions.	my $WorkQueue = Queue->new(); # Queue of pending transactions.
my $ClientQueue = Queue->new(); # Queue of clients causing xactinos.
my $ConnectionCount = 0;	my $ConnectionCount = 0;
my $IdleSeconds = 0; # Number of seconds idle.	my $IdleSeconds = 0; # Number of seconds idle.
	my $Status = ""; # Current status string.
	my $RecentLogEntry = "";
	my $ConnectionRetries=5; # Number of connection retries allowed.
	my $ConnectionRetriesLeft=5; # Number of connection retries remaining.

#	#
# This disconnected socket makes posible a bit more regular	# The hash below gives the HTML format for log messages
# code when processing delayed requests:	# given a severity.
#	#
my $NullSocket = IO::Socket->new();	my %LogFormats;

	$LogFormats{"CRITICAL"} = "<font color=red>CRITICAL: %s</font>";
	$LogFormats{"SUCCESS"} = "<font color=green>SUCCESS: %s</font>";
	$LogFormats{"INFO"} = "<font color=yellow>INFO: %s</font>";
	$LogFormats{"WARNING"} = "<font color=blue>WARNING: %s</font>";
	$LogFormats{"DEFAULT"} = " %s ";



	=pod

	=head2 LogPerm

	Makes an entry into the permanent log file.

	=cut
	sub LogPerm {
	my $message=shift;
	my $execdir=$perlvar{'lonDaemons'};
	my $now=time;
	my $local=localtime($now);
	my $fh=IO::File->new(">>$execdir/logs/lonnet.perm.log");
	print $fh "$now:$message:$local\n";
	}

	=pod

	=head2 Log

	Logs a message to the log file.
	Parameters:

	=item severity

	One of CRITICAL, WARNING, INFO, SUCCESS used to select the
	format string used to format the message. if the severity is
	not a defined severity the Default format string is used.

	=item message

	The base message. In addtion to the format string, the message
	will be appended to a string containing the name of our remote
	host and the time will be formatted into the message.

	=cut

	sub Log {
	my $severity = shift;
	my $message = shift;

	if(!$LogFormats{$severity}) {
	$severity = "DEFAULT";
	}

	my $format = $LogFormats{$severity};

	# Put the window dressing in in front of the message format:

	my $now = time;
	my $local = localtime($now);
	my $finalformat = "$local ($$) [$RemoteHost] [$Status] ";
	my $finalformat = $finalformat.$format."\n";

	# open the file and put the result.

	my $execdir = $perlvar{'lonDaemons'};
	my $fh = IO::File->new(">>$execdir/logs/lonc.log");
	my $msg = sprintf($finalformat, $message);
	$RecentLogEntry = $msg;
	print $fh $msg;


	}

#

=pod	=pod

Line 127 sub GetPeername {	Line 281 sub GetPeername {
my $peerip;	my $peerip;
if($AdrFamily == AF_INET) {	if($AdrFamily == AF_INET) {
($peerport, $peerip) = sockaddr_in($peer);	($peerport, $peerip) = sockaddr_in($peer);
my $peername = gethostbyaddr($iaddr, $AdrFamily);	my $peername = gethostbyaddr($peerip, $AdrFamily);
return $peername;	return $peername;
} elsif ($AdrFamily == AF_UNIX) {	} elsif ($AdrFamily == AF_UNIX) {
my $peerfile;	my $peerfile;
Line 148 sub Debug {	Line 302 sub Debug {
my $level = shift;	my $level = shift;
my $message = shift;	my $message = shift;
if ($level <= $DebugLevel) {	if ($level <= $DebugLevel) {
print $message." host = ".$RemoteHost."\n";	Log("INFO", "-Debug- $message host = $RemoteHost");
}	}
}	}

Line 165 sub SocketDump {	Line 319 sub SocketDump {
=head2 ShowStatus	=head2 ShowStatus

Place some text as our pid status.	Place some text as our pid status.
	and as what we return in a SIGUSR1

=cut	=cut
sub ShowStatus {	sub ShowStatus {
my $status = shift;	my $state = shift;
$0 = "lonc: ".$status;	my $now = time;
	my $local = localtime($now);
	$Status = $local.": ".$state;
	$0='lonc: '.$state.' '.$local;
	}

	=pod

	=head 2 SocketTimeout

	Called when an action on the socket times out. The socket is
	destroyed and any active transaction is failed.


	=cut
	sub SocketTimeout {
	my $Socket = shift;

	KillSocket($Socket); # A transaction timeout also counts as
	# a connection failure:
	$ConnectionRetriesLeft--;
}	}

=pod	=pod
Line 183 Invoked each timer tick.	Line 358 Invoked each timer tick.

sub Tick {	sub Tick {
my $client;	my $client;
ShowStatus(GetServerHost()." Connection count: ".$ConnectionCount);	if($ConnectionRetriesLeft > 0) {
Debug(6, "Tick");	ShowStatus(GetServerHost()." Connection count: ".$ConnectionCount
Debug(6, " Current connection count: ".$ConnectionCount);	." Retries remaining: ".$ConnectionRetriesLeft);
foreach $client (keys %ActiveClients) {	} else {
Debug(7, " Have client: with id: ".$ActiveClients{$client});	ShowStatus(GetServerHost()." >> DEAD <<");
}	}
# Is it time to prune connection count:	# Is it time to prune connection count:

Line 196 sub Tick {	Line 371 sub Tick {
($WorkQueue->Count() == 0)) { # Idle connections and nothing to do?	($WorkQueue->Count() == 0)) { # Idle connections and nothing to do?
$IdleSeconds++;	$IdleSeconds++;
if($IdleSeconds > $IdleTimeout) { # Prune a connection...	if($IdleSeconds > $IdleTimeout) { # Prune a connection...
$Socket = $IdleConnections->pop();	my $Socket = $IdleConnections->pop();
KillSocket($Socket);	KillSocket($Socket);
}	}
} else {	} else {
$IdleSeconds = 0; # Reset idle count if not idle.	$IdleSeconds = 0; # Reset idle count if not idle.
}	}
	#
	# For each inflight transaction, tick down its timeout counter.
	#
	foreach my $item (keys %ActiveTransactions) {
	my $Socket = $ActiveTransactions{$item}->getServer();
	$Socket->Tick();
	}
# Do we have work in the queue, but no connections to service them?	# Do we have work in the queue, but no connections to service them?
# If so, try to make some new connections to get things going again.	# If so, try to make some new connections to get things going again.
#	#

my $Requests = $WorkQueue->Count();	my $Requests = $WorkQueue->Count();
if (($ConnectionCount == 0) && ($Requests > 0)) {	if (($ConnectionCount == 0) && ($Requests > 0)) {
my $Connections = ($Requests <= $MaxConnectionCount) ?	if ($ConnectionRetriesLeft > 0) {
$Requests : $MaxConnectionCount;	my $Connections = ($Requests <= $MaxConnectionCount) ?
Debug(1,"Work but no connections, starting ".$Connections." of them");	$Requests : $MaxConnectionCount;
for ($i =0; $i < $Connections; $i++) {	Debug(1,"Work but no connections, start ".$Connections." of them");
MakeLondConnection();	my $successCount = 0;
	for (my $i =0; $i < $Connections; $i++) {
	$successCount += MakeLondConnection();
	}
	if($successCount == 0) { # All connections failed:
	Debug(1,"Work in queue failed to make any connectiouns\n");
	EmptyQueue(); # Fail pending transactions with con_lost.
	}
	} else {
	ShowStatus(GetServerHost()." >>> DEAD!!! <<<");
	Debug(1,"Work in queue, but gave up on connections..flushing\n");
	EmptyQueue(); # Connections can't be established.
}	}

}	}
Line 257 long enough, it will be shut down and re	Line 449 long enough, it will be shut down and re

sub ServerToIdle {	sub ServerToIdle {
my $Socket = shift; # Get the socket.	my $Socket = shift; # Get the socket.
	delete($ActiveTransactions{$Socket}); # Server has no transaction

&Debug(6, "Server to idle");	&Debug(6, "Server to idle");

# If there's work to do, start the transaction:	# If there's work to do, start the transaction:

$reqdata = $WorkQueue->dequeue();	my $reqdata = $WorkQueue->dequeue(); # This is a LondTransaction
Debug(9, "Queue gave request data: ".$reqdata);
unless($reqdata eq undef) {	unless($reqdata eq undef) {
my $unixSocket = $ClientQueue->dequeue();	Debug(9, "Queue gave request data: ".$reqdata->getRequest());
&Debug(6, "Starting new work request");	&StartRequest($Socket, $reqdata);
&Debug(7, "Request: ".$reqdata);

&StartRequest($Socket, $unixSocket, $reqdata);
} else {	} else {

# There's no work waiting, so push the server to idle list.	# There's no work waiting, so push the server to idle list.
&Debug(8, "No new work requests, server connection going idle");	&Debug(8, "No new work requests, server connection going idle");
delete($ActiveTransactions{$Socket});
$IdleConnections->push($Socket);	$IdleConnections->push($Socket);
}	}
}	}
Line 338 sub ClientWritable {	Line 527 sub ClientWritable {

} else { # Partial string sent.	} else { # Partial string sent.
$Watcher->data(substr($Data, $result));	$Watcher->data(substr($Data, $result));
	if($result == 0) { # client hung up on us!!
	Log("INFO", "lonc pipe client hung up on us!");
	$Watcher->cancel;
	$Socket->shutdown(2);
	$Socket->close();
	}
}	}

} else { # Error of some sort...	} else { # Error of some sort...
Line 378 Parameters:	Line 573 Parameters:
Socket on which the lond transaction occured. This is a	Socket on which the lond transaction occured. This is a
LondConnection. The data received is in the TransactionReply member.	LondConnection. The data received is in the TransactionReply member.

=item Client	=item Transaction

Unix domain socket open on the ultimate client.	The transaction that is being completed.

=cut	=cut

sub CompleteTransaction {	sub CompleteTransaction {
&Debug(6,"Complete transaction");	&Debug(6,"Complete transaction");
my $Socket = shift;	my $Socket = shift;
my $Client = shift;	my $Transaction = shift;

my $data = $Socket->GetReply(); # Data to send.	if (!$Transaction->isDeferred()) { # Normal transaction
StartClientReply($Client, $data);	my $data = $Socket->GetReply(); # Data to send.
	StartClientReply($Transaction, $data);
	} else { # Delete deferred transaction file.
	Log("SUCCESS", "A delayed transaction was completed");
	LogPerm("S:$Transaction->getClient() :".$Transaction->getRequest());
	unlink $Transaction->getFile();
	}
}	}
=pod	=pod
=head1 StartClientReply	=head1 StartClientReply

Initiates a reply to a client where the reply data is a parameter.	Initiates a reply to a client where the reply data is a parameter.

	=head2 parameters:

	=item Transaction

	The transaction for which we are responding to the client.

	=item data

	The data to send to apached client.

=cut	=cut
sub StartClientReply {	sub StartClientReply {
my $Client = shift;	my $Transaction = shift;
my $data = shift;	my $data = shift;


	my $Client = $Transaction->getClient();

&Debug(8," Reply was: ".$data);	&Debug(8," Reply was: ".$data);
my $Serial = $ActiveClients{$Client};	my $Serial = $ActiveClients{$Client};
my $desc = sprintf("Connection to lonc client %d",	my $desc = sprintf("Connection to lonc client %d",
Line 417 sub StartClientReply {	Line 631 sub StartClientReply {
=head2 FailTransaction	=head2 FailTransaction

Finishes a transaction with failure because the associated lond socket	Finishes a transaction with failure because the associated lond socket
disconnected. It is up to our client to retry if desired.	disconnected. There are two possibilities:
	- The transaction is deferred: in which case we just quietly
	delete the transaction since there is no client connection.
	- The transaction is 'live' in which case we initiate the sending
	of "con_lost" to the client.

	Deleting the transaction means killing it from the
	%ActiveTransactions hash.

Parameters:	Parameters:

=item client	=item client

The UNIX domain socket open on our client.	The LondTransaction we are failing.

=cut	=cut

sub FailTransaction {	sub FailTransaction {
my $client = shift;	my $transaction = shift;
	Log("WARNING", "Failing transaction ".$transaction->getRequest());
StartClientReply($client, "con_lost");	Debug(1, "Failing transaction: ".$transaction->getRequest());
	if (!$transaction->isDeferred()) { # If the transaction is deferred we'll get to it.
	my $client = $transaction->getClient();
	Debug(1," Replying con_lost to ".$transaction->getRequest());
	StartClientReply($transaction, "con_lost\n");
	}
	if($ConnectionRetriesLeft <= 0) {
	Log("CRITICAL", "Host marked dead: ".GetServerHost());
	}

}	}

=pod	=pod
=head1 EmptyQueue	=head1 EmptyQueue

Fails all items in the work queue with con_lost.	Fails all items in the work queue with con_lost.
	Note that each item in the work queue is a transaction.

=cut	=cut
sub EmptyQueue {	sub EmptyQueue {
	$ConnectionRetriesLeft--; # Counts as connection failure too.
while($WorkQueue->Count()) {	while($WorkQueue->Count()) {
my $request = $WorkQUeue->dequeue(); # Just to help it become empty.	my $request = $WorkQueue->dequeue(); # This is a transaction
my $client = $ClientQueue->dequeue(); # Need to con_lost this guy.	FailTransaction($request);
FailTransaction($client);	}
	}

	=pod

	=head2 CloseAllLondConnections

	Close all connections open on lond prior to exit e.g.

	=cut
	sub CloseAllLondConnections {
	foreach my $Socket (keys %ActiveConnections) {
	KillSocket($Socket);
}	}
}	}
	=cut

=pod	=pod

Line 471 nonzero if we are allowed to create a ne	Line 717 nonzero if we are allowed to create a ne
sub KillSocket {	sub KillSocket {
my $Socket = shift;	my $Socket = shift;

# If the socket came from the active connection set, delete it.	Log("WARNING", "Shutting down a socket");
# otherwise it came from the idle set and has already been destroyed:	$Socket->Shutdown();

	# If the socket came from the active connection set,
	# delete its transaction... note that FailTransaction should
	# already have been called!!!
	# otherwise it came from the idle set.
	#

if(exists($ActiveTransactions{$Socket})) {	if(exists($ActiveTransactions{$Socket})) {
delete ($ActiveTransactions{$Socket});	delete ($ActiveTransactions{$Socket});
Line 486 sub KillSocket {	Line 738 sub KillSocket {
# work queue, the work all gets failed with con_lost.	# work queue, the work all gets failed with con_lost.
#	#
if($ConnectionCount == 0) {	if($ConnectionCount == 0) {
EmptyQueue;	EmptyQueue();
}	}
}	}

Line 549 transaction is in progress, the socket a	Line 801 transaction is in progress, the socket a
=cut	=cut

sub LondReadable {	sub LondReadable {

my $Event = shift;	my $Event = shift;
my $Watcher = $Event->w;	my $Watcher = $Event->w;
my $Socket = $Watcher->data;	my $Socket = $Watcher->data;
my $client = undef;	my $client = undef;

	&Debug(6,"LondReadable called state = ".$Socket->GetState());


my $State = $Socket->GetState(); # All action depends on the state.	my $State = $Socket->GetState(); # All action depends on the state.

&Debug(6,"LondReadable called state = ".$State);
SocketDump(6, $Socket);	SocketDump(6, $Socket);
	my $status = $Socket->Readable();

	&Debug(2, "Socket->Readable returned: $status");

if($Socket->Readable() != 0) {	if($status != 0) {
# bad return from socket read. Currently this means that	# bad return from socket read. Currently this means that
# The socket has become disconnected. We fail the transaction.	# The socket has become disconnected. We fail the transaction.

	Log("WARNING",
	"Lond connection lost.");
if(exists($ActiveTransactions{$Socket})) {	if(exists($ActiveTransactions{$Socket})) {
Debug(3,"Lond connection lost failing transaction");
FailTransaction($ActiveTransactions{$Socket});	FailTransaction($ActiveTransactions{$Socket});
}	}
$Watcher->cancel();	$Watcher->cancel();
KillSocket($Socket);	KillSocket($Socket);
	$ConnectionRetriesLeft--; # Counts as connection failure
return;	return;
}	}
SocketDump(6,$Socket);	SocketDump(6,$Socket);
Line 585 sub LondReadable {	Line 844 sub LondReadable {
# in the connection takes care of setting that up. Just	# in the connection takes care of setting that up. Just
# need to transition to writable:	# need to transition to writable:

$Watcher->poll("w");
$Watcher->cb(\&LondWritable);	$Watcher->cb(\&LondWritable);
	$Watcher->poll("w");

} elsif ($State eq "ChallengeReplied") {	} elsif ($State eq "ChallengeReplied") {

Line 595 sub LondReadable {	Line 854 sub LondReadable {
# The ok was received. Now we need to request the key	# The ok was received. Now we need to request the key
# That requires us to be writable:	# That requires us to be writable:

$Watcher->poll("w");
$Watcher->cb(\&LondWritable);	$Watcher->cb(\&LondWritable);
	$Watcher->poll("w");

} elsif ($State eq "ReceivingKey") {	} elsif ($State eq "ReceivingKey") {

} elsif ($State eq "Idle") {	} elsif ($State eq "Idle") {
# If necessary, complete a transaction and then go into the	# If necessary, complete a transaction and then go into the
# idle queue.	# idle queue.
	# Note that a trasition to idle indicates a live lond
	# on the other end so reset the connection retries.
	#
	$ConnectionRetriesLeft = $ConnectionRetries; # success resets the count
	$Watcher->cancel();
if(exists($ActiveTransactions{$Socket})) {	if(exists($ActiveTransactions{$Socket})) {
Debug(8,"Completing transaction!!");	Debug(8,"Completing transaction!!");
CompleteTransaction($Socket,	CompleteTransaction($Socket,
$ActiveTransactions{$Socket});	$ActiveTransactions{$Socket});
	} else {
	Log("SUCCESS", "Connection ".$ConnectionCount." to "
	.$RemoteHost." now ready for action");
}	}
$Watcher->cancel();
ServerToIdle($Socket); # Next work unit or idle.	ServerToIdle($Socket); # Next work unit or idle.

} elsif ($State eq "SendingRequest") {	} elsif ($State eq "SendingRequest") {
Line 692 is the socket on which to return a reply	Line 958 is the socket on which to return a reply
sub LondWritable {	sub LondWritable {
my $Event = shift;	my $Event = shift;
my $Watcher = $Event->w;	my $Watcher = $Event->w;
my @data = $Watcher->data;	my $Socket = $Watcher->data;
Debug(6,"LondWritable State = ".$State." data has ".@data." elts.\n");	my $State = $Socket->GetState();

my $Socket = $data[0]; # I know there's at least a socket.	Debug(6,"LondWritable State = ".$State."\n");


# Figure out what to do depending on the state of the socket:	# Figure out what to do depending on the state of the socket:


my $State = $Socket->GetState();


SocketDump(6,$Socket);	SocketDump(6,$Socket);
Line 710 sub LondWritable {	Line 976 sub LondWritable {
if ($Socket->Writable() != 0) {	if ($Socket->Writable() != 0) {
# The write resulted in an error.	# The write resulted in an error.
# We'll treat this as if the socket got disconnected:	# We'll treat this as if the socket got disconnected:
	Log("WARNING", "Connection to ".$RemoteHost.
	" has been disconnected");
	FailTransaction($ActiveTransactions{$Socket});
$Watcher->cancel();	$Watcher->cancel();
KillSocket($Socket);	KillSocket($Socket);
return;	return;
Line 723 sub LondWritable {	Line 991 sub LondWritable {
# Now that init was sent, we switch	# Now that init was sent, we switch
# to watching for readability:	# to watching for readability:

$Watcher->poll("r");
$Watcher->cb(\&LondReadable);	$Watcher->cb(\&LondReadable);
	$Watcher->poll("r");

} elsif ($State eq "ChallengeReceived") {	} elsif ($State eq "ChallengeReceived") {
# We received the challenge, now we	# We received the challenge, now we
Line 742 sub LondWritable {	Line 1010 sub LondWritable {
# The echo was sent back, so we switch	# The echo was sent back, so we switch
# to watching readability.	# to watching readability.

$Watcher->poll("r");
$Watcher->cb(\&LondReadable);	$Watcher->cb(\&LondReadable);
	$Watcher->poll("r");

} elsif ($State eq "RequestingKey") {	} elsif ($State eq "RequestingKey") {
# At this time we're requesting the key.	# At this time we're requesting the key.
Line 763 sub LondWritable {	Line 1031 sub LondWritable {
# Now we need to wait for the key	# Now we need to wait for the key
# to come back from the peer:	# to come back from the peer:

$Watcher->poll("r");
$Watcher->cb(\&LondReadable);	$Watcher->cb(\&LondReadable);
	$Watcher->poll("r");

} elsif ($State eq "SendingRequest") {	} elsif ($State eq "SendingRequest") {
# At this time we are sending a request to the	# At this time we are sending a request to the
Line 786 sub LondWritable {	Line 1054 sub LondWritable {
# The send has completed. Wait for the	# The send has completed. Wait for the
# data to come in for a reply.	# data to come in for a reply.
Debug(8,"Writable sent request/receiving reply");	Debug(8,"Writable sent request/receiving reply");
$Watcher->poll("r");
$Watcher->cb(\&LondReadable);	$Watcher->cb(\&LondReadable);
	$Watcher->poll("r");

} else {	} else {
# Control only passes here on an error:	# Control only passes here on an error:
Line 803 sub LondWritable {	Line 1071 sub LondWritable {

=cut	=cut
sub QueueDelayed {	sub QueueDelayed {
	Debug(3,"QueueDelayed called");

my $path = "$perlvar{'lonSockDir'}/delayed";	my $path = "$perlvar{'lonSockDir'}/delayed";

	Debug(4, "Delayed path: ".$path);
opendir(DIRHANDLE, $path);	opendir(DIRHANDLE, $path);
@alldelayed = grep /\.$RemoteHost$/, readdir DIRHANDLE;
	my @alldelayed = grep /\.$RemoteHost$/, readdir DIRHANDLE;
closedir(DIRHANDLE);	closedir(DIRHANDLE);
my $dfname;	my $dfname;
my $reqfile	my $reqfile;
foreach $reqfile (sort @alldelayed) {	foreach $dfname (sort @alldelayed) {
$reqfile = $path/$reqfile;	$reqfile = "$path/$dfname";
	Debug(4, "queueing ".$reqfile);
my $Handle = IO::File->new($reqfile);	my $Handle = IO::File->new($reqfile);
my $cmd = <$Handle>;	my $cmd = <$Handle>;
chomp($cmd);	chomp $cmd; # There may or may not be a newline...
QueueTransaction($NullSocket, $cmd);	$cmd = $cmd."\n"; # now for sure there's exactly one newline.
	my $Transaction = LondTransaction->new($cmd);
	$Transaction->SetDeferred($reqfile);
	QueueTransaction($Transaction);
}	}

}	}
Line 839 sub MakeLondConnection {	Line 1116 sub MakeLondConnection {
&GetServerPort());	&GetServerPort());

if($Connection == undef) { # Needs to be more robust later.	if($Connection == undef) { # Needs to be more robust later.
Debug(0,"Failed to make a connection with lond.");	Log("CRITICAL","Failed to make a connection with lond.");
	$ConnectionRetriesLeft--;
	return 0; # Failure.
} else {	} else {

# The connection needs to have writability	# The connection needs to have writability
# monitored in order to send the init sequence	# monitored in order to send the init sequence
# that starts the whole authentication/key	# that starts the whole authentication/key
Line 853 sub MakeLondConnection {	Line 1133 sub MakeLondConnection {
&Debug(9,"MakeLondConnection got socket: ".$Socket);	&Debug(9,"MakeLondConnection got socket: ".$Socket);
}	}

	$Connection->SetTimeoutCallback(\&SocketTimeout);
$event = Event->io(fd => $Socket,
	my $event = Event->io(fd => $Socket,
poll => 'w',	poll => 'w',
cb => \&LondWritable,	cb => \&LondWritable,
data => ($Connection, undef),	data => $Connection,
desc => 'Connection to lond server');	desc => 'Connection to lond server');
$ActiveConnections{$Connection} = $event;	$ActiveConnections{$Connection} = $event;

$ConnectionCount++;	$ConnectionCount++;
	Debug(4, "Connection count = ".$ConnectionCount);
if($ConnectionCount == 1) { # First Connection:	if($ConnectionCount == 1) { # First Connection:
QueueDelayed;	QueueDelayed;
}	}
	Log("SUCESS", "Created connection ".$ConnectionCount
	." to host ".GetServerHost());
	return 1; # Return success.
}	}

}	}
Line 896 The text of the request to send.	Line 1181 The text of the request to send.

sub StartRequest {	sub StartRequest {
my $Lond = shift;	my $Lond = shift;
my $Client = shift;	my $Request = shift; # This is a LondTransaction.
my $Request = shift;

Debug(6, "StartRequest: ".$Request);	Debug(6, "StartRequest: ".$Request->getRequest());

my $Socket = $Lond->GetSocket();	my $Socket = $Lond->GetSocket();

$ActiveTransactions{$Lond} = $Client; # Socket to relay to client.	$Request->Activate($Lond);
	$ActiveTransactions{$Lond} = $Request;

$Lond->InitiateTransaction($Request);	$Lond->InitiateTransaction($Request->getRequest());
$event = Event->io(fd => $Lond->GetSocket(),	my $event = Event->io(fd => $Socket,
poll => "w",	poll => "w",
cb => \&LondWritable,	cb => \&LondWritable,
data => $Lond,	data => $Lond,
Line 937 data to send to the lond.	Line 1222 data to send to the lond.
=cut	=cut

sub QueueTransaction {	sub QueueTransaction {
my $requestSocket = shift;
my $requestData = shift;

Debug(6,"QueueTransaction: ".$requestData);	my $requestData = shift; # This is a LondTransaction.
	my $cmd = $requestData->getRequest();

	Debug(6,"QueueTransaction: ".$cmd);

my $LondSocket = $IdleConnections->pop();	my $LondSocket = $IdleConnections->pop();
if(!defined $LondSocket) { # Need to queue request.	if(!defined $LondSocket) { # Need to queue request.
Debug(8,"Must queue...");	Debug(8,"Must queue...");
$ClientQueue->enqueue($requestSocket);
$WorkQueue->enqueue($requestData);	$WorkQueue->enqueue($requestData);
if($ConnectionCount < $MaxConnectionCount) {	if($ConnectionCount < $MaxConnectionCount) {
Debug(4,"Starting additional lond connection");	if($ConnectionRetriesLeft > 0) {
MakeLondConnection();	Debug(4,"Starting additional lond connection");
	if(MakeLondConnection() == 0) {
	EmptyQueue(); # Fail transactions, can't make connection.
	}
	} else {
	ShowStatus(GetServerHost()." >>> DEAD !!!! <<<");
	EmptyQueue(); # It's worse than that ... he's dead Jim.
	}
}	}
} else { # Can start the request:	} else { # Can start the request:
Debug(8,"Can start...");	Debug(8,"Can start...");
StartRequest($LondSocket, $requestSocket, $requestData);	StartRequest($LondSocket, $requestData);
}	}
}	}

Line 962 sub QueueTransaction {	Line 1254 sub QueueTransaction {
=pod	=pod

=head2 ClientRequest	=head2 ClientRequest

Callback that is called when data can be read from the UNIX domain	Callback that is called when data can be read from the UNIX domain
socket connecting us with an apache server process.	socket connecting us with an apache server process.

Line 987 sub ClientRequest {	Line 1278 sub ClientRequest {
close($socket);	close($socket);
$watcher->cancel();	$watcher->cancel();
delete($ActiveClients{$socket});	delete($ActiveClients{$socket});
	return;
}	}
Debug(8,"Data: ".$data." this read: ".$thisread);	Debug(8,"Data: ".$data." this read: ".$thisread);
$data = $data.$thisread; # Append new data.	$data = $data.$thisread; # Append new data.
$watcher->data($data);	$watcher->data($data);
if($data =~ /(.*\n)/) { # Request entirely read.	if($data =~ /(.*\n)/) { # Request entirely read.
	if($data eq "close_connection_exit\n") {
	Log("CRITICAL",
	"Request Close Connection ... exiting");
	CloseAllLondConnections();
	exit;
	}
Debug(8, "Complete transaction received: ".$data);	Debug(8, "Complete transaction received: ".$data);
QueueTransaction($socket, $data);	my $Transaction = LondTransaction->new($data);
	$Transaction->SetClient($socket);
	QueueTransaction($Transaction);
$watcher->cancel(); # Done looking for input data.	$watcher->cancel(); # Done looking for input data.
}	}

Line 1055 Returns the host whose lond we talk with	Line 1355 Returns the host whose lond we talk with

=cut	=cut

sub GetServerHost { # Stub - get this from config.	sub GetServerHost {
return $RemoteHost; # Setup by the fork.	return $RemoteHost; # Setup by the fork.
}	}

Line 1067 Returns the lond port number.	Line 1367 Returns the lond port number.

=cut	=cut

sub GetServerPort { # Stub - get this from config.	sub GetServerPort {
return $perlvar{londPort};	return $perlvar{londPort};
}	}

Line 1088 sub SetupLoncListener {	Line 1388 sub SetupLoncListener {
my $socket;	my $socket;
my $SocketName = GetLoncSocketPath();	my $SocketName = GetLoncSocketPath();
unlink($SocketName);	unlink($SocketName);
unless ($socket = IO::Socket::UNIX->new(Local => $SocketName,	unless ($socket =IO::Socket::UNIX->new(Local => $SocketName,
Listen => 10,	Listen => 10,
Type => SOCK_STREAM)) {	Type => SOCK_STREAM)) {
die "Failed to create a lonc listner socket";	die "Failed to create a lonc listner socket";
Line 1099 sub SetupLoncListener {	Line 1399 sub SetupLoncListener {
fd => $socket);	fd => $socket);
}	}

	=pod

	=head2 ChildStatus

	Child USR1 signal handler to report the most recent status
	into the status file.

	We also use this to reset the retries count in order to allow the
	client to retry connections with a previously dead server.
	=cut
	sub ChildStatus {
	my $event = shift;
	my $watcher = $event->w;

	Debug(2, "Reporting child status because : ".$watcher->data);
	my $docdir = $perlvar{'lonDocRoot'};
	my $fh = IO::File->new(">>$docdir/lon-status/loncstatus.txt");
	print $fh $$."\t".$RemoteHost."\t".$Status."\t".
	$RecentLogEntry."\n";
	$ConnectionRetriesLeft = $ConnectionRetries;
	}

=pod	=pod

	=head2 SignalledToDeath

	Called in response to a signal that causes a chid process to die.

	=cut


	sub SignalledToDeath {
	my $event = shift;
	my $watcher= $event->w;

	Debug(2,"Signalled to death! via ".$watcher->data);
	my ($signal) = $watcher->data;
	chomp($signal);
	Log("CRITICAL", "Abnormal exit. Child $$ for $RemoteHost "
	."died through "."\"$signal\"");
	LogPerm("F:lonc: $$ on $RemoteHost signalled to death: "
	."\"$signal\"");
	exit 0;

	}

	=head2 ToggleDebug

	This sub toggles trace debugging on and off.

	=cut

	sub ToggleDebug {
	my $Current = $DebugLevel;
	$DebugLevel = $NextDebugLevel;
	$NextDebugLevel = $Current;

	Log("SUCCESS", "New debugging level for $RemoteHost now $DebugLevel");

	}

=head2 ChildProcess	=head2 ChildProcess

This sub implements a child process for a single lonc daemon.	This sub implements a child process for a single lonc daemon.
Line 1109 This sub implements a child process for	Line 1468 This sub implements a child process for

sub ChildProcess {	sub ChildProcess {

print "Loncnew\n";

# For now turn off signals.	#
	# Signals must be handled by the Event framework...
$SIG{QUIT} = IGNORE;	#
$SIG{HUP} = IGNORE;
$SIG{USR1} = IGNORE;	Event->signal(signal => "QUIT",
$SIG{INT} = IGNORE;	cb => \&SignalledToDeath,
$SIG{CHLD} = IGNORE;	data => "QUIT");
$SIG{__DIE__} = IGNORE;	Event->signal(signal => "HUP",
	cb => \&ChildStatus,
	data => "HUP");
	Event->signal(signal => "USR1",
	cb => \&ChildStatus,
	data => "USR1");
	Event->signal(signal => "INT",
	cb => \&ToggleDebug,
	data => "INT");

SetupTimer();	SetupTimer();

Line 1130 sub ChildProcess {	Line 1496 sub ChildProcess {

# Setup the initial server connection:	# Setup the initial server connection:

&MakeLondConnection();	# &MakeLondConnection(); // let first work requirest do it.


if($ConnectionCount == 0) {
Debug(1,"Could not make initial connection..\n");
Debug(1,"Will retry when there's work to do\n");
}
Debug(9,"Entering event loop");	Debug(9,"Entering event loop");
my $ret = Event::loop(); # Start the main event loop.	my $ret = Event::loop(); # Start the main event loop.

Line 1146 sub ChildProcess {	Line 1509 sub ChildProcess {
# Create a new child for host passed in:	# Create a new child for host passed in:

sub CreateChild {	sub CreateChild {
	my $sigset = POSIX::SigSet->new(SIGINT);
	sigprocmask(SIG_BLOCK, $sigset);
my $host = shift;	my $host = shift;
$RemoteHost = $host;	$RemoteHost = $host;
Debug(3, "Forking off child for ".$RemoteHost);	Log("CRITICAL", "Forking server for ".$host);
sleep(5);	my $pid = fork;
$pid = fork;
if($pid) { # Parent	if($pid) { # Parent
	$RemoteHost = "Parent";
$ChildHash{$pid} = $RemoteHost;	$ChildHash{$pid} = $RemoteHost;
	sigprocmask(SIG_UNBLOCK, $sigset);

} else { # child.	} else { # child.
ShowStatus("Connected to ".$RemoteHost);	ShowStatus("Connected to ".$RemoteHost);
ChildProcess;	$SIG{INT} = 'DEFAULT';
	sigprocmask(SIG_UNBLOCK, $sigset);
	ChildProcess; # Does not return.
}	}

}	}
Line 1177 sub CreateChild {	Line 1546 sub CreateChild {



ShowStatus("Parent writing pid file:");
$execdir = $perlvar{'lonDaemons'};
open (PIDSAVE, ">$execdir/logs/lonc.pid");
print PIDSAVE "$$\n";
close(PIDSAVE);

ShowStatus("Forming new session");	ShowStatus("Forming new session");
my $childpid = fork;	my $childpid = fork;
Line 1189 if ($childpid != 0) {	Line 1554 if ($childpid != 0) {
sleep 4; # Give child a chacne to break to	sleep 4; # Give child a chacne to break to
exit 0; # a new sesion.	exit 0; # a new sesion.
}	}
	#
	# Write my pid into the pid file so I can be located
	#

	ShowStatus("Parent writing pid file:");
	my $execdir = $perlvar{'lonDaemons'};
	open (PIDSAVE, ">$execdir/logs/lonc.pid");
	print PIDSAVE "$$\n";
	close(PIDSAVE);



if (POSIX::setsid() < 0) {	if (POSIX::setsid() < 0) {
print "Could not create new session\n";	print "Could not create new session\n";
Line 1197 if (POSIX::setsid() < 0) {	Line 1573 if (POSIX::setsid() < 0) {

ShowStatus("Forking node servers");	ShowStatus("Forking node servers");

	Log("CRITICAL", "--------------- Starting children ---------------");

my $HostIterator = LondConnection::GetHostIterator;	my $HostIterator = LondConnection::GetHostIterator;
while (! $HostIterator->end()) {	while (! $HostIterator->end()) {

$hostentryref = $HostIterator->get();	my $hostentryref = $HostIterator->get();
CreateChild($hostentryref->[0]);	CreateChild($hostentryref->[0]);
$HostIterator->next();	$HostIterator->next();
}	}
	$RemoteHost = "Parent Server";

# Maintain the population:	# Maintain the population:

ShowStatus("Parent keeping the flock");	ShowStatus("Parent keeping the flock");

	#
	# Set up parent signals:
	#

	$SIG{INT} = \&Terminate;
	$SIG{TERM} = \&Terminate;
	$SIG{HUP} = \&Restart;
	$SIG{USR1} = \&CheckKids;

while(1) {	while(1) {
$deadchild = wait();	my $deadchild = wait();
if(exists $ChildHash{$deadchild}) { # need to restart.	if(exists $ChildHash{$deadchild}) { # need to restart.
$deadhost = $ChildHash{$deadchild};	my $deadhost = $ChildHash{$deadchild};
delete($ChildHash{$deadchild});	delete($ChildHash{$deadchild});
Debug(4,"Lost child pid= ".$deadchild.	Log("WARNING","Lost child pid= ".$deadchild.
"Connected to host ".$deadhost);	"Connected to host ".$deadhost);
	Log("INFO", "Restarting child procesing ".$deadhost);
CreateChild($deadhost);	CreateChild($deadhost);
}	}
}	}



	=pod

	=head1 CheckKids

	Since kids do not die as easily in this implementation
	as the previous one, there is no need to restart the
	dead ones (all dead kids get restarted when they die!!)
	The only thing this function does is to pass USR1 to the
	kids so that they report their status.

	=cut

	sub CheckKids {
	Debug(2, "Checking status of children");
	my $docdir = $perlvar{'lonDocRoot'};
	my $fh = IO::File->new(">$docdir/lon-status/loncstatus.txt");
	my $now=time;
	my $local=localtime($now);
	print $fh "LONC status $local - parent $$ \n\n";
	foreach my $pid (keys %ChildHash) {
	Debug(2, "Sending USR1 -> $pid");
	kill 'USR1' => $pid; # Tell Child to report status.
	sleep 1; # Wait so file doesn't intermix.
	}
	}

	=pod

	=head1 Restart

	Signal handler for HUP... all children are killed and
	we self restart. This is an el-cheapo way to re read
	the config file.

	=cut

	sub Restart {
	&KillThemAll; # First kill all the children.
	Log("CRITICAL", "Restarting");
	my $execdir = $perlvar{'lonDaemons'};
	unlink("$execdir/logs/lonc.pid");
	exec("$execdir/lonc");
	}

	=pod

	=head1 KillThemAll

	Signal handler that kills all children by sending them a
	SIGHUP. Responds to sigint and sigterm.

	=cut

	sub KillThemAll {
	Debug(2, "Kill them all!!");
	local($SIG{CHLD}) = 'IGNORE'; # Our children >will< die.
	foreach my $pid (keys %ChildHash) {
	my $serving = $ChildHash{$pid};
	Debug(2, "Killing lonc for $serving pid = $pid");
	ShowStatus("Killing lonc for $serving pid = $pid");
	Log("CRITICAL", "Killing lonc for $serving pid = $pid");
	kill 'QUIT' => $pid;
	delete($ChildHash{$pid});
	}
	my $execdir = $perlvar{'lonDaemons'};
	unlink("$execdir/logs/lonc.pid");

	}

	=pod

	=head1 Terminate

	Terminate the system.

	=cut

	sub Terminate {
	KillThemAll;
	Log("CRITICAL","Master process exiting");
	exit 0;

	}
	=pod

=head1 Theory	=head1 Theory

The event class is used to build this as a single process with an	The event class is used to build this as a single process with an

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>

Removed from v.1.6
changed lines
	Added in v.1.23