--- loncom/loncron 2010/08/17 22:22:00 1.86 +++ loncom/loncron 2015/09/20 18:31:21 1.103 @@ -2,7 +2,7 @@ # Housekeeping program, started by cron, loncontrol and loncron.pl # -# $Id: loncron,v 1.86 2010/08/17 22:22:00 raeburn Exp $ +# $Id: loncron,v 1.103 2015/09/20 18:31:21 raeburn Exp $ # # Copyright Michigan State University Board of Trustees # @@ -32,6 +32,8 @@ use strict; use lib '/home/httpd/lib/perl/'; use LONCAPA::Configuration; +use LONCAPA::Checksumming; +use LONCAPA; use Apache::lonnet; use Apache::loncommon; @@ -155,12 +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)) { @@ -269,12 +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); @@ -288,9 +285,11 @@ sub start_logging { &log($fh,(< + + LON Status Report $perlvar{'lonHostID'} + @@ -301,6 +300,7 @@ sub start_logging {
  • Machine Information
  • Temporary Files
  • Session Tokens
  • +
  • WebDAV Session Tokens
  • httpd
  • lonsql
  • lond
  • @@ -331,15 +331,20 @@ ENDHEADERS "".$role. "".&Apache::lonnet::hostname($id)."\n"); } - &log($fh,"

    Spare Hosts

    \n"); return $fh; } @@ -485,7 +490,7 @@ sub clean_lonIDs { my $since=$now-$mtime; if ($since>$perlvar{'lonExpire'}) { $cleaned++; - &log($fh,"Unlinking $fname
    "); + &log($fh,"Unlinking $fname
    "); unlink("$fname"); } else { $active++; @@ -495,6 +500,32 @@ sub clean_lonIDs { &log($fh,"

    $active open session(s)

    "); } +# ------------------------------------------------ clean out webDAV Session IDs +sub clean_webDAV_sessionIDs { + my ($fh)=@_; + if ($perlvar{'lonRole'} eq 'library') { + &log($fh,'

    WebDAV Session Tokens

    '); + my $cleaned=0; + my $active=0; + my $now = time; + if (-d $perlvar{'lonDAVsessDir'}) { + while (my $fname=<$perlvar{'lonDAVsessDir'}/*>) { + my @stats = stat($fname); + my $since=$now-$stats[9]; + if ($since>$perlvar{'lonExpire'}) { + $cleaned++; + &log($fh,"Unlinking $fname
    "); + unlink("$fname"); + } else { + $active++; + } + } + &log($fh,"

    Cleaned up ".$cleaned." stale webDAV session token(s).

    "); + &log($fh,"

    $active open webDAV session(s)

    "); + } + } +} + # ----------------------------------------------------------- clean out sockets sub clean_sockets { my ($fh)=@_; @@ -514,21 +545,13 @@ sub clean_sockets { # ----------------------------------------------------------------------- httpd sub check_httpd_logs { my ($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($fh); } @@ -537,7 +560,7 @@ sub check_httpd_logs { sub rotate_lonnet_logs { my ($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 $line=) { 
    @@ -581,7 +604,7 @@ sub rotate_other_logs {
     sub test_connections {
         my ($fh)=@_;
         &log($fh,'

    Connections

    '); - print "testing connections\n"; + print "Testing connections.\n"; &log($fh,""); my ($good,$bad)=(0,0); my %hostname = &Apache::lonnet::all_hostnames(); @@ -615,7 +638,7 @@ sub test_connections { sub check_delayed_msg { my ($fh)=@_; &log($fh,'

    Delayed Messages

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

    Scanning Permanent Log

    '); @@ -626,7 +649,7 @@ 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') { $unsend--; } @@ -634,7 +657,9 @@ sub check_delayed_msg { } &log($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
    ");
    @@ -648,11 +673,28 @@ 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'.
    "); + } } } @@ -665,7 +707,7 @@ 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; } @@ -686,43 +728,94 @@ sub log_simplestatus { } sub write_loncaparevs { - if (open(my $fh,">$perlvar{'lonTabDir'}/loncaparevs.tab")) { - my %hostname = &Apache::lonnet::all_hostnames(); - foreach my $id (sort(keys(%hostname))) { - if ($id ne '') { - my $loncaparev = &Apache::lonnet::get_server_loncaparev('',$id,1,'loncron'); - if ($loncaparev =~ /^[\w.\-]+$/) { - print $fh $id.':'.$loncaparev."\n"; - } + 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"; } } - close($fh); + } + if ($output) { + if (open(my $fh,">$perlvar{'lonTabDir'}/loncaparevs.tab")) { + print $fh $output; + close($fh); + &Apache::lonnet::load_loncaparevs(); + } } return; } sub write_serverhomeIDs { - if (open(my $fh,">$perlvar{'lonTabDir'}/serverhomeIDs.tab")) { - my %name_to_host = &Apache::lonnet::all_names(); - foreach my $name (sort(keys(%name_to_host))) { - if ($name ne '') { - if (ref($name_to_host{$name}) eq 'ARRAY') { - my $serverhomeID = &Apache::lonnet::get_server_homeID($name,1,'loncron'); - if ($serverhomeID ne '') { - print $fh $name.':'.$serverhomeID."\n"; - } else { - print $fh $name.':'.$name_to_host{$name}->[0]."\n"; - } + 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{$name}->[0]."\n"; } } } - close($fh); + } + 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 $defdom = $perlvar{'lonDefDomain'}; my $origmail = $perlvar{'lonAdmEMail'}; my $emailto = &Apache::loncommon::build_recipient_list(undef, @@ -730,17 +823,40 @@ sub send_mail { if ($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; } # -------------------------------------------- Force reload of host information - &Apache::lonnet::load_hosts_tab(1); - &Apache::lonnet::load_domain_tab(1); - &Apache::lonnet::get_iphost(1); + my $nomemcache; + if ($justcheckdaemons) { + $nomemcache=1; + my $memcachepidfile="$perlvar{'lonDaemons'}/logs/memcached.pid"; + my $memcachepid; + if (-e $memcachepidfile) { + my $memfh=IO::File->new($memcachepidfile); + $memcachepid=<$memfh>; + chomp($memcachepid); + if ($memcachepid =~ /^\d+$/ && kill 0 => $memcachepid) { + undef($nomemcache); + } + } + } + &Apache::lonnet::load_hosts_tab(1,$nomemcache); + &Apache::lonnet::load_domain_tab(1,$nomemcache); + &Apache::lonnet::get_iphost(1,$nomemcache); # ----------------------------------------- Force firewall update for lond port @@ -821,9 +951,12 @@ sub main () { print $fh "$key\n"; } close($fh); - my $execpath = $perlvar{'lonDaemons'}.'/lciptables'; - system("$execpath $tmpfile"); - unlink($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); } } @@ -841,6 +974,7 @@ sub main () { &log_machine_info($fh); &clean_tmp($fh); &clean_lonIDs($fh); + &clean_webDAV_sessionIDs($fh); &check_httpd_logs($fh); &rotate_lonnet_logs($fh); &rotate_other_logs($fh); @@ -866,22 +1000,13 @@ sub main () { &check_delayed_msg($fh); &finish_logging($fh); &log_simplestatus(); - - if ($totalcount>200 && !$noemail) { &send_mail(); } - } - if (!$justcheckconnections && !$justreload) { &write_loncaparevs(); &write_serverhomeIDs(); + &write_checksums(); + if ($totalcount>200 && !$noemail) { &send_mail(); } } } &main(); 1; - - - - - - -