--- loncom/loncnew 2003/06/13 02:38:43 1.9 +++ loncom/loncnew 2003/06/24 02:46:04 1.10 @@ -2,7 +2,7 @@ # The LearningOnline Network with CAPA # lonc maintains the connections to remote computers # -# $Id: loncnew,v 1.9 2003/06/13 02:38:43 foxr Exp $ +# $Id: loncnew,v 1.10 2003/06/24 02:46:04 foxr Exp $ # # Copyright Michigan State University Board of Trustees # @@ -46,6 +46,11 @@ # Change log: # $Log: loncnew,v $ +# 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 6the client will permanently give up on dead servers. +# # Revision 1.9 2003/06/13 02:38:43 foxr # Add logging in 'expected format' # @@ -120,7 +125,8 @@ my $WorkQueue = Queue->new(); # Qu my $ConnectionCount = 0; my $IdleSeconds = 0; # Number of seconds idle. my $Status = ""; # Current status string. - +my $ConnectionRetries=5; # Number of connection retries allowed. +my $ConnectionRetriesLeft=5; # Number of connection retries remaining. # # The hash below gives the HTML format for log messages @@ -134,7 +140,23 @@ $LogFormats{"INFO"} = "new(">>$execdir/logs/lonnet.perm.log"); + print $fh "$now:$message:$local\n"; +} =pod @@ -181,6 +203,7 @@ sub Log { my $msg = sprintf($finalformat, $message); print $fh $msg; + } @@ -238,13 +261,15 @@ sub SocketDump { =head2 ShowStatus Place some text as our pid status. + and as what we return in a SIGUSR1 =cut sub ShowStatus { - my $status = shift; - $0 = "lonc: ".$status; - $Status = $status; # Make available for logging. - + my $state = shift; + my $now = time; + my $local = localtime($now); + $Status = $local.": ".$state; + $0='lonc: '.$state.' '.$local; } =pod @@ -283,12 +308,17 @@ sub Tick { # my $Requests = $WorkQueue->Count(); - if (($ConnectionCount == 0) && ($Requests > 0)) { - my $Connections = ($Requests <= $MaxConnectionCount) ? - $Requests : $MaxConnectionCount; - Debug(1,"Work but no connections, starting ".$Connections." of them"); - for ($i =0; $i < $Connections; $i++) { - MakeLondConnection(); + if (($ConnectionCount == 0) && ($Requests > 0)) { + if ($ConnectionRetriesLeft > 0) { + my $Connections = ($Requests <= $MaxConnectionCount) ? + $Requests : $MaxConnectionCount; + Debug(1,"Work but no connections, start ".$Connections." of them"); + for ($i =0; $i < $Connections; $i++) { + MakeLondConnection(); + } + } else { + Debug(1,"Work in queue, but gave up on connections..flushing\n"); + EmptyQueue(); # Connections can't be established. } } @@ -466,6 +496,7 @@ sub CompleteTransaction { StartClientReply($Transaction, $data); } else { # Delete deferred transaction file. Log("SUCCESS", "A delayed transaction was completed"); + LogPerm("S:$Client:".$Transaction->getRequest()); unlink $Transaction->getFile(); } } @@ -525,13 +556,12 @@ Parameters: sub FailTransaction { my $transaction = shift; - my $Lond = $transaction->getServer(); - if (!$client->isDeferred()) { # If the transaction is deferred we'll get to it. + Debug(1, "Failing transaction: ".$transaction->getRequest()); + if (!$transaction->isDeferred()) { # If the transaction is deferred we'll get to it. my $client = $transcation->getClient(); - StartClientReply($client, "con_lost"); + Debug(1," Replying con_lost to ".$transaction->getRequest()); + StartClientReply($client, "con_lost\n"); } -# not needed, done elsewhere if active. -# delete $ActiveTransactions{$Lond}; } @@ -544,7 +574,7 @@ sub FailTransaction { =cut sub EmptyQueue { while($WorkQueue->Count()) { - my $request = $Workqueue->dequeue(); # This is a transaction + my $request = $WorkQueue->dequeue(); # This is a transaction FailTransaction($request); } } @@ -978,7 +1008,10 @@ sub MakeLondConnection { if($Connection == undef) { # Needs to be more robust later. Log("CRITICAL","Failed to make a connection with lond."); + $ConnectionRetriesLeft--; + return 0; # Failure. } else { + $ConnectionRetriesLeft = $ConnectionRetries; # success resets the count # The connection needs to have writability # monitored in order to send the init sequence # that starts the whole authentication/key @@ -1006,6 +1039,7 @@ sub MakeLondConnection { } Log("SUCESS", "Created connection ".$ConnectionCount ." to host ".GetServerHost()); + return 1; # Return success. } } @@ -1103,7 +1137,6 @@ sub QueueTransaction { =pod =head2 ClientRequest - Callback that is called when data can be read from the UNIX domain socket connecting us with an apache server process. @@ -1128,12 +1161,13 @@ sub ClientRequest { close($socket); $watcher->cancel(); delete($ActiveClients{$socket}); + return; } Debug(8,"Data: ".$data." this read: ".$thisread); $data = $data.$thisread; # Append new data. $watcher->data($data); if($data =~ /(.*\n)/) { # Request entirely read. - if($data == "close_connection_exit\n") { + if($data eq "close_connection_exit\n") { Log("CRITICAL", "Request Close Connection ... exiting"); CloseAllLondConnections(); @@ -1250,6 +1284,24 @@ sub SetupLoncListener { =pod +=head2 SignalledToDeath + +Called in response to a signal that causes a chid process to die. + +=cut + +=pod + +sub SignalledToDeath { + my ($signal) = @_; + chomp($signal); + Log("CRITICAL", "Abnormal exit. Child $$ for $RemoteHost " + ."died through "."\"$signal\""); + LogPerm("F:lonc: $$ on $RemoteHost signalled to death: " + ."\"$signal\""); + die("Signal abnormal end"); + +} =head2 ChildProcess This sub implements a child process for a single lonc daemon. @@ -1261,12 +1313,12 @@ sub ChildProcess { # For now turn off signals. - $SIG{QUIT} = IGNORE; + $SIG{QUIT} = \&SignalledToDeath; $SIG{HUP} = IGNORE; $SIG{USR1} = IGNORE; $SIG{INT} = IGNORE; $SIG{CHLD} = IGNORE; - $SIG{__DIE__} = IGNORE; + $SIG{__DIE__} = \&SignalledToDeath; SetupTimer(); @@ -1278,12 +1330,9 @@ sub ChildProcess { # Setup the initial server connection: - &MakeLondConnection(); + # &MakeLondConnection(); // let first work requirest do it. + - if($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"); my $ret = Event::loop(); # Start the main event loop. @@ -1363,6 +1412,12 @@ while (! $HostIterator->end()) { ShowStatus("Parent keeping the flock"); +# +# Set up parent signals: +# +$SIG{INT} = &KillThemAll; +$SIG{TERM} = &KillThemAll; + while(1) { $deadchild = wait(); if(exists $ChildHash{$deadchild}) { # need to restart. @@ -1374,6 +1429,8 @@ while(1) { CreateChild($deadhost); } } +sub KillThemAll { +} =head1 Theory