--- loncom/loncnew 2003/10/27 10:09:21 1.30
+++ loncom/loncnew 2004/01/05 09:29:36 1.38
@@ -2,7 +2,7 @@
# The LearningOnline Network with CAPA
# lonc maintains the connections to remote computers
#
-# $Id: loncnew,v 1.30 2003/10/27 10:09:21 foxr Exp $
+# $Id: loncnew,v 1.38 2004/01/05 09:29:36 foxr Exp $
#
# Copyright Michigan State University Board of Trustees
#
@@ -35,118 +35,16 @@
# - Add ability to create/negotiate lond connections (done).
# - Add general logic for dispatching requests and timeouts. (done).
# - Add support for the lonc/lond requests. (done).
-# - Add logging/status monitoring.
-# - Add Signal handling - HUP restarts. USR1 status report.
+# - Add logging/status monitoring. (done)
+# - Add Signal handling - HUP restarts. USR1 status report. (done)
# - Add Configuration file I/O (done).
-# - Add management/status request interface.
+# - Add management/status request interface. (done)
# - Add deferred request capability. (done)
-# - Detect transmission timeouts.
+# - Detect transmission timeouts. (done)
#
-# Change log:
-# $Log: loncnew,v $
-# Revision 1.30 2003/10/27 10:09:21 foxr
-# Tighten up a few compares to eq and flip a few debug levels around... nothing
-# critical
-#
-# Revision 1.29 2003/10/21 14:24:42 foxr
-# Fix little typo that may explain growth of connections
-#
-# Revision 1.28 2003/10/14 15:36:21 albertel
-# - making it easier to run loncnew,
-# /etc/init.d/loncontrol startnew
-# /etc/init.d/loncontrol restartnew
-# will now start loncnew in place of lonc
-#
-# Revision 1.27 2003/10/07 11:23:03 foxr
-# Installed and tested code to process reinit in parent server.
-#
-# Revision 1.26 2003/09/30 11:11:17 foxr
-# Add book-keeping hashes to support the re-init procedure.
-#
-# Revision 1.25 2003/09/23 11:22:14 foxr
-# Tested ability to receive sigusr2 This is now logged and must be
-# properly implemented as a re-read of hosts and re-init of appropriate
-# children.
-#
-# Revision 1.24 2003/09/16 09:46:42 foxr
-# Added skeletal infrastructure to support SIGUSR2 update hosts request.
-#
-# 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/foxr/newloncapa/types";
use Event qw(:DEFAULT );
use POSIX qw(:signal_h);
use POSIX;
@@ -312,7 +210,6 @@ sub GetPeername {
return $peerfile;
}
}
-#----------------------------- Timer management ------------------------
=pod
=head2 Debug
@@ -364,11 +261,14 @@ sub ShowStatus {
=cut
sub SocketTimeout {
my $Socket = shift;
-
+ Log("WARNING", "A socket timeout was detected");
+ Debug(0, " SocketTimeout called: ");
+ $Socket->Dump();
KillSocket($Socket); # A transaction timeout also counts as
# a connection failure:
$ConnectionRetriesLeft--;
}
+#----------------------------- Timer management ------------------------
=pod
@@ -403,9 +303,13 @@ sub Tick {
#
# For each inflight transaction, tick down its timeout counter.
#
- foreach my $item (keys %ActiveTransactions) {
- my $Socket = $ActiveTransactions{$item}->getServer();
- $Socket->Tick();
+
+ foreach my $item (keys %ActiveConnections) {
+ my $State = $ActiveConnections{$item}->data->GetState();
+ if ($State ne 'Idle') {
+ Debug(5,"Ticking Socket $State $item");
+ $ActiveConnections{$item}->data->Tick();
+ }
}
# 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.
@@ -452,7 +356,7 @@ Trigger disconnections of idle sockets.
sub SetupTimer {
Debug(6, "SetupTimer");
- Event->timer(interval => 1, debug => 1, cb => \&Tick );
+ Event->timer(interval => 1, cb => \&Tick );
}
=pod
@@ -753,9 +657,9 @@ sub KillSocket {
}
if(exists($ActiveConnections{$Socket})) {
delete($ActiveConnections{$Socket});
+ $ConnectionCount--;
+ if ($ConnectionCount < 0) { $ConnectionCount = 0; }
}
- $ConnectionCount--;
-
# If the connection count has gone to zero and there is work in the
# work queue, the work all gets failed with con_lost.
#
@@ -1440,6 +1344,20 @@ sub ChildStatus {
my $fh = IO::File->new(">>$docdir/lon-status/loncstatus.txt");
print $fh $$."\t".$RemoteHost."\t".$Status."\t".
$RecentLogEntry."\n";
+ #
+ # Write out information about each of the connections:
+ #
+ print $fh "Active connection statuses: \n";
+ my $i = 1;
+ print STDERR "================================= Socket Status Dump:\n";
+ foreach my $item (keys %ActiveConnections) {
+ my $Socket = $ActiveConnections{$item}->data;
+ my $state = $Socket->GetState();
+ print $fh "Connection $i State: $state\n";
+ print STDERR "---------------------- Connection $i \n";
+ $Socket->Dump();
+ $i++;
+ }
$ConnectionRetriesLeft = $ConnectionRetries;
}
@@ -1598,6 +1516,7 @@ ShowStatus("Forking node servers");
Log("CRITICAL", "--------------- Starting children ---------------");
+LondConnection::ReadConfig; # Read standard config files.
my $HostIterator = LondConnection::GetHostIterator;
while (! $HostIterator->end()) {