--- loncom/loncron 2008/11/04 21:06:21 1.77 +++ loncom/loncron 2014/12/16 18:18:44 1.101 @@ -2,7 +2,7 @@ # Housekeeping program, started by cron, loncontrol and loncron.pl # -# $Id: loncron,v 1.77 2008/11/04 21:06:21 www Exp $ +# $Id: loncron,v 1.101 2014/12/16 18:18:44 raeburn Exp $ # # Copyright Michigan State University Board of Trustees # @@ -32,7 +32,10 @@ use strict; use lib '/home/httpd/lib/perl/'; use LONCAPA::Configuration; +use LONCAPA::Checksumming; +use LONCAPA; use Apache::lonnet; +use Apache::loncommon; use IO::File; use IO::Socket; @@ -154,12 +157,12 @@ sub checkon_daemon { $errors++; my $kadaemon=$daemon; if ($kadaemon eq 'lonmemcached') { $kadaemon='memcached'; } - &log($fh,'
Killall '.$daemon.': '. + &log($fh,'
Killall '.$daemon.': '. `killall $kadaemon 2>&1`.' - '); sleep 1; &log($fh,unlink($pidfile).' - '. `killall -9 $kadaemon 2>&1`. - '
'); + '

'); &log($fh,"

$daemon not running, trying to start

"); if (&start_daemon($fh,$daemon,$pidfile,$args)) { @@ -268,12 +271,7 @@ sub log_machine_info { &log($fh,"

distprobe

"); &log($fh,"
");
-    open(DSH,"$perlvar{'lonDaemons'}/distprobe |");
-    while (my $line=) { 
-	&log($fh,&encode_entities($line,'<>&"')); 
-	$psproc++;
-    }
-    close(DSH);
+    &log($fh,&encode_entities(&LONCAPA::distro(),'<>&"'));
     &log($fh,"
"); &errout($fh); @@ -287,9 +285,11 @@ sub start_logging { &log($fh,(< + + LON Status Report $perlvar{'lonHostID'} + @@ -330,15 +330,20 @@ ENDHEADERS "".$role. "".&Apache::lonnet::hostname($id)."\n"); } - &log($fh,"

Spare Hosts

\n"); return $fh; } @@ -346,33 +351,127 @@ ENDHEADERS sub clean_tmp { my ($fh)=@_; &log($fh,'

Temporary Files

'); - my $cleaned=0; - my $old=0; - while (my $fname=<$perlvar{'lonDaemons'}/tmp/*>) { - my ($dev,$ino,$mode,$nlink, - $uid,$gid,$rdev,$size, - $atime,$mtime,$ctime, - $blksize,$blocks)=stat($fname); - my $now=time; - my $since=$now-$mtime; - if ($since>$perlvar{'lonExpire'}) { - my $line=''; - if (open(PROBE,$fname)) { - $line=; - close(PROBE); - } - unless ($line=~/^CHECKOUTTOKEN\&/) { - $cleaned++; - unlink("$fname"); - } else { - if ($since>365*$perlvar{'lonExpire'}) { - $cleaned++; - unlink("$fname"); - } else { $old++; } - } - } + my ($cleaned,$old,$removed) = (0,0,0); + my %errors = ( + dir => [], + file => [], + failopen => [], + ); + my %error_titles = ( + dir => 'failed to remove empty directory:', + file => 'failed to unlike stale file', + failopen => 'failed to open file or directory' + ); + ($cleaned,$old,$removed) = &recursive_clean_tmp('',$cleaned,$old,$removed,\%errors); + &log($fh,"Cleaned up: ".$cleaned." files; removed: $removed empty directories; (found: $old old checkout tokens)"); + foreach my $key (sort(keys(%errors))) { + if (ref($errors{$key}) eq 'ARRAY') { + if (@{$errors{$key}} > 0) { + &log($fh,"Error during cleanup ($error_titles{$key}):
  • ". + join('
  • ',@{$errors{$key}}).'

'); + } + } + } +} + +sub recursive_clean_tmp { + my ($subdir,$cleaned,$old,$removed,$errors) = @_; + my $base = "$perlvar{'lonDaemons'}/tmp"; + my $path = $base; + next if ($subdir =~ m{\.\./}); + next unless (ref($errors) eq 'HASH'); + unless ($subdir eq '') { + $path .= '/'.$subdir; + } + if (opendir(my $dh,"$path")) { + while (my $file = readdir($dh)) { + next if ($file =~ /^\.\.?$/); + my $fname = "$path/$file"; + if (-d $fname) { + my $innerdir; + if ($subdir eq '') { + $innerdir = $file; + } else { + $innerdir = $subdir.'/'.$file; + } + ($cleaned,$old,$removed) = + &recursive_clean_tmp($innerdir,$cleaned,$old,$removed,$errors); + my @doms = &Apache::lonnet::current_machine_domains(); + + if (open(my $dirhandle,$fname)) { + unless (($innerdir eq 'helprequests') || + (($innerdir =~ /^addcourse/) && ($innerdir !~ m{/\d+$}))) { + my @contents = grep {!/^\.\.?$/} readdir($dirhandle); + join('&&',@contents)."\n"; + if (scalar(grep {!/^\.\.?$/} readdir($dirhandle)) == 0) { + closedir($dirhandle); + if ($fname =~ m{^\Q$perlvar{'lonDaemons'}\E/tmp/}) { + if (rmdir($fname)) { + $removed ++; + } elsif (ref($errors->{dir}) eq 'ARRAY') { + push(@{$errors->{dir}},$fname); + } + } + } + } else { + closedir($dirhandle); + } + } + } else { + my ($dev,$ino,$mode,$nlink, + $uid,$gid,$rdev,$size, + $atime,$mtime,$ctime, + $blksize,$blocks)=stat($fname); + my $now=time; + my $since=$now-$mtime; + if ($since>$perlvar{'lonExpire'}) { + if ($subdir eq '') { + my $line=''; + if ($fname =~ /\.db$/) { + if (unlink($fname)) { + $cleaned++; + } elsif (ref($errors->{file}) eq 'ARRAY') { + push(@{$errors->{file}},$fname); + } + } elsif (open(PROBE,$fname)) { + my $line=''; + $line=; + close(PROBE); + if ($line=~/^CHECKOUTTOKEN\&/) { + if ($since>365*$perlvar{'lonExpire'}) { + if (unlink($fname)) { + $cleaned++; + } elsif (ref($errors->{file}) eq 'ARRAY') { + push(@{$errors->{file}},$fname); + } + } else { + $old++; + } + } else { + if (unlink($fname)) { + $cleaned++; + } elsif (ref($errors->{file}) eq 'ARRAY') { + push(@{$errors->{file}},$fname); + } + } + } elsif (ref($errors->{failopen}) eq 'ARRAY') { + push(@{$errors->{failopen}},$fname); + } + } else { + if (unlink($fname)) { + $cleaned++; + } elsif (ref($errors->{file}) eq 'ARRAY') { + push(@{$errors->{file}},$fname); + } + } + } + } + } + closedir($dh); + } elsif (ref($errors->{failopen}) eq 'ARRAY') { + push(@{$errors->{failopen}},$path); } - &log($fh,"Cleaned up ".$cleaned." files (".$old." old checkout tokens)."); + return ($cleaned,$old,$removed); } # ------------------------------------------------------------ clean out lonIDs @@ -390,7 +489,7 @@ sub clean_lonIDs { my $since=$now-$mtime; if ($since>$perlvar{'lonExpire'}) { $cleaned++; - &log($fh,"Unlinking $fname
"); + &log($fh,"Unlinking $fname
"); unlink("$fname"); } else { $active++; @@ -407,7 +506,7 @@ sub clean_sockets { opendir(SOCKETS,$perlvar{'lonSockDir'}); while (my $fname=readdir(SOCKETS)) { next if (-d $fname - || $fname=~/(mysqlsock|maximasock|\Q$perlvar{'lonSockDir'}\E)/); + || $fname=~/(mysqlsock|maximasock|rsock|\Q$perlvar{'lonSockDir'}\E)/); $cleaned++; &log($fh,"Unlinking $fname
"); unlink("/home/httpd/sockets/$fname"); @@ -419,21 +518,13 @@ sub clean_sockets { # ----------------------------------------------------------------------- httpd sub check_httpd_logs { my ($fh)=@_; - &log($fh,'

httpd

Access Log

');
-    
-    open (DFH,"tail -n25 /etc/httpd/logs/access_log|");
-    while (my $line=) { &log($fh,&encode_entities($line,'<>&"')) };
-    close (DFH);
-	
-    &log($fh,"

Error Log

");
-	
-    open (DFH,"tail -n25 /etc/httpd/logs/error_log|");
-    while (my $line=) { 
-	&log($fh,"$line");
-	if ($line=~/\[error\]/) { $notices++; } 
+    if (open(PIPE,"./lchttpdlogs|")) {
+        while (my $line=) {
+            &log($fh,$line);
+            if ($line=~/\[error\]/) { $notices++; }
+        }
+        close(PIPE);
     }
-    close (DFH);
-    &log($fh,"
"); &errout($fh); } @@ -442,7 +533,7 @@ sub check_httpd_logs { sub rotate_lonnet_logs { my ($fh)=@_; &log($fh,'

lonnet

Temp Log

');
-    print "checking logs\n";
+    print "Checking logs.\n";
     if (-e "$perlvar{'lonDaemons'}/logs/lonnet.log"){
 	open (DFH,"tail -n50 $perlvar{'lonDaemons'}/logs/lonnet.log|");
 	while (my $line=) { 
@@ -469,19 +560,24 @@ sub rotate_lonnet_logs {
 
 sub rotate_other_logs {
     my ($fh) = @_;
-    my $fname="$perlvar{'lonDaemons'}/logs/autoenroll.log";
-    &rotate_logfile($fname,$fh,'Auto Enroll log');
-    $fname="$perlvar{'lonDaemons'}/logs/autocreate.log";
-    &rotate_logfile($fname,$fh,'Create Course log');
-    $fname="$perlvar{'lonDaemons'}/logs/searchcat.log";
-    &rotate_logfile($fname,$fh,'Search Cataloguing log');
+    my %logs = (
+                  autoenroll          => 'Auto Enroll log',
+                  autocreate          => 'Create Course log',
+                  searchcat           => 'Search Cataloguing log',
+                  autoupdate          => 'Auto Update log',
+                  refreshcourseids_db => 'Refresh CourseIDs db log',
+               );
+    foreach my $item (keys(%logs)) {
+        my $fname=$perlvar{'lonDaemons'}.'/logs/'.$item.'.log';
+        &rotate_logfile($fname,$fh,$logs{$item});
+    }
 }
 
 # ----------------------------------------------------------------- Connections
 sub test_connections {
     my ($fh)=@_;
     &log($fh,'

Connections

'); - print "testing connections\n"; + print "Testing connections.\n"; &log($fh,""); my ($good,$bad)=(0,0); my %hostname = &Apache::lonnet::all_hostnames(); @@ -515,7 +611,7 @@ sub test_connections { sub check_delayed_msg { my ($fh)=@_; &log($fh,'

Delayed Messages

'); - print "checking buffers\n"; + print "Checking buffers.\n"; &log($fh,'

Scanning Permanent Log

'); @@ -526,7 +622,7 @@ sub check_delayed_msg { my ($time,$sdf,$dserv,$dcmd)=split(/:/,$line); if ($sdf eq 'F') { my $local=localtime($time); - &log($fh,"Failed: $time, $dserv, $dcmd
"); + &log($fh,"Failed: $time, $dserv, $dcmd
"); $warnings++; } if ($sdf eq 'S') { $unsend--; } @@ -534,7 +630,9 @@ sub check_delayed_msg { } &log($fh,"

Total unsend messages: $unsend

\n"); - $warnings=$warnings+5*$unsend; + if ($unsend > 0) { + $warnings=$warnings+5*$unsend; + } if ($unsend) { $simplestatus{'unsend'}=$unsend; } &log($fh,"

Outgoing Buffer

\n
");
@@ -548,11 +646,28 @@ sub check_delayed_msg {
     }
     &log($fh,"
\n"); close (DFH); + my %hostname = &Apache::lonnet::all_hostnames(); + my $numhosts = scalar(keys(%hostname)); # pong to all servers that have delayed messages # this will trigger a reverse connection, which should flush the buffers - foreach my $tryserver (keys %servers) { - my $answer=&Apache::lonnet::reply("pong",$tryserver); - &log($fh,"Pong to $tryserver: $answer
"); + foreach my $tryserver (sort(keys(%servers))) { + if ($hostname{$tryserver} || !$numhosts) { + my $answer; + eval { + local $SIG{ ALRM } = sub { die "TIMEOUT" }; + alarm(20); + $answer = &Apache::lonnet::reply("pong",$tryserver); + alarm(0); + }; + if ($@ && $@ =~ m/TIMEOUT/) { + &log($fh,"Attempted pong to $tryserver timed out
"); + print "Time out while contacting: $tryserver for pong.\n"; + } else { + &log($fh,"Pong to $tryserver: $answer
"); + } + } else { + &log($fh,"$tryserver has delayed messages, but is not part of the cluster -- skipping 'Pong'.
"); + } } } @@ -565,7 +680,7 @@ sub finish_logging { my $now=time; my $date=localtime($now); &log($fh,"
$date ($now)\n"); - print "lon-status webpage updated\n"; + print "lon-status webpage updated.\n"; $fh->close(); if ($errors) { $simplestatus{'errors'}=$errors; } @@ -585,23 +700,136 @@ sub log_simplestatus { $sfh->close(); } +sub write_loncaparevs { + print "Retrieving LON-CAPA version information.\n"; + my %hostname = &Apache::lonnet::all_hostnames(); + my $output; + foreach my $id (sort(keys(%hostname))) { + if ($id ne '') { + my $loncaparev; + eval { + local $SIG{ ALRM } = sub { die "TIMEOUT" }; + alarm(10); + $loncaparev = + &Apache::lonnet::get_server_loncaparev('',$id,1,'loncron'); + alarm(0); + }; + if ($@ && $@ =~ m/TIMEOUT/) { + print "Time out while contacting lonHost: $id for version.\n"; + } + if ($loncaparev =~ /^[\w.\-]+$/) { + $output .= $id.':'.$loncaparev."\n"; + } + } + } + if ($output) { + if (open(my $fh,">$perlvar{'lonTabDir'}/loncaparevs.tab")) { + print $fh $output; + close($fh); + &Apache::lonnet::load_loncaparevs(); + } + } + return; +} + +sub write_serverhomeIDs { + print "Retrieving LON-CAPA lonHostID information.\n"; + my %name_to_host = &Apache::lonnet::all_names(); + my $output; + foreach my $name (sort(keys(%name_to_host))) { + if ($name ne '') { + if (ref($name_to_host{$name}) eq 'ARRAY') { + my $serverhomeID; + eval { + local $SIG{ ALRM } = sub { die "TIMEOUT" }; + alarm(10); + $serverhomeID = + &Apache::lonnet::get_server_homeID($name,1,'loncron'); + alarm(0); + }; + if ($@ && $@ =~ m/TIMEOUT/) { + print "Time out while contacting server: $name\n"; + } + if ($serverhomeID ne '') { + $output .= $name.':'.$serverhomeID."\n"; + } else { + $output .= $name.':'.$name_to_host{$name}->[0]."\n"; + } + } + } + } + if ($output) { + if (open(my $fh,">$perlvar{'lonTabDir'}/serverhomeIDs.tab")) { + print $fh $output; + close($fh); + &Apache::lonnet::load_serverhomeIDs(); + } + } + return; +} + +sub write_checksums { + my $distro = &LONCAPA::distro(); + if ($distro) { + print "Retrieving file version and checksumming.\n"; + my $numchksums = 0; + my ($chksumsref,$versionsref) = + &LONCAPA::Checksumming::get_checksums($distro,$perlvar{'lonDaemons'}, + $perlvar{'lonLib'}, + $perlvar{'lonIncludes'}, + $perlvar{'lonTabDir'}); + if (ref($chksumsref) eq 'HASH') { + $numchksums = scalar(keys(%{$chksumsref})); + } + print "File version retrieved and checksumming completed for $numchksums files.\n"; + } else { + print "File version retrieval and checksumming skipped - could not determine Linux distro.\n"; + } + return; +} + sub send_mail { - print "sending mail\n"; - my $emailto="$perlvar{'lonAdmEMail'}"; + my $defdom = $perlvar{'lonDefDomain'}; + my $origmail = $perlvar{'lonAdmEMail'}; + my $emailto = &Apache::loncommon::build_recipient_list(undef, + 'lonstatusmail',$defdom,$origmail); if ($totalcount>2500) { $emailto.=",$perlvar{'lonSysEMail'}"; } - my $subj="LON: $perlvar{'lonHostID'} E:$errors W:$warnings N:$notices"; - - my $result=system("metasend -b -S 4000000 -t $emailto -s '$subj' -f $statusdir/index.html -m text/html >& /dev/null"); - if ($result != 0) { - $result=system("mail -s '$subj' $emailto < $statusdir/index.html"); + my $from; + my $hostname=`/bin/hostname`; + chop($hostname); + $hostname=~s/[^\w\.]//g; + if ($hostname) { + $from = 'www@'.$hostname; + } + my $subj="LON: $perlvar{'lonHostID'} E:$errors W:$warnings N:$notices"; + my $loncronmail = "To: $emailto\n". + "From: $from\n". + "Subject: ".$subj."\n". + "Content-type: text/html\; charset=UTF-8\n". + "MIME-Version: 1.0\n\n"; + if (open(my $fh,"<$statusdir/index.html")) { + while (<$fh>) { + $loncronmail .= $_; + } + close($fh); + } else { + $loncronmail .= "Failed to read from http://$hostname/lon-status/index.html\n"; + } + $loncronmail .= "\n\n"; + if (open(my $mailh, "|/usr/lib/sendmail -oi -t -odb")) { + print $mailh $loncronmail; + close($mailh); + print "Sending mail.\n"; + } else { + print "Sending mail failed.\n"; } } sub usage { print(< /dev/null"); exit 1; } @@ -670,6 +898,27 @@ sub main () { &Apache::lonnet::load_domain_tab(1); &Apache::lonnet::get_iphost(1); +# ----------------------------------------- Force firewall update for lond port + + if ((!$justcheckdaemons) && (!$justreload)) { + my $now = time; + my $tmpfile = $perlvar{'lonDaemons'}.'/tmp/lciptables_iphost_'. + $now.$$.int(rand(10000)); + if (open(my $fh,">$tmpfile")) { + my %iphosts = &Apache::lonnet::get_iphost(); + foreach my $key (keys(%iphosts)) { + print $fh "$key\n"; + } + close($fh); + if (&LONCAPA::try_to_lock('/tmp/lock_lciptables')) { + my $execpath = $perlvar{'lonDaemons'}.'/lciptables'; + system("$execpath $tmpfile"); + unlink('/tmp/lock_lciptables'); # Remove the lock file. + } + unlink($tmpfile); + } + } + # ---------------------------------------------------------------- Start report $errors=0; @@ -696,6 +945,7 @@ sub main () { } &checkon_daemon($fh,'lonc',40000,'USR1'); &checkon_daemon($fh,'lonmaxima',40000); + &checkon_daemon($fh,'lonr',40000); } if ($justreload) { &checkon_daemon($fh,'lond',40000,'USR2'); @@ -708,7 +958,9 @@ sub main () { &check_delayed_msg($fh); &finish_logging($fh); &log_simplestatus(); - + &write_loncaparevs(); + &write_serverhomeIDs(); + &write_checksums(); if ($totalcount>200 && !$noemail) { &send_mail(); } } } @@ -716,10 +968,3 @@ sub main () { &main(); 1; - - - - - - -