Add -D | --dns-resolv command line option to replace ip adresses by their DNS name...

diff --git a/README b/README
index 847b8ff45d81a17f042f04250e4243dd77353be9..3dba42b1480417662394310416df818f178fee20 100644 (file)
--- a/README
+++ b/README
@@ -23,6 +23,8 @@ SYNOPSIS
         -c | --dbclient host   : only report on entries for the given client host.
         -C | --nocomment       : remove comments like /* ... */ from queries.
         -d | --dbname database : only report on entries for the given database.
+        -D | --dns-resolv      : client ip adresses are replaced by their DNS name.
+                                 Be warned that this can really slow down pgBadger.
         -e | --end datetime    : end date/time for the data to be parsed in log.
         -f | --format logtype  : possible values: syslog,stderr,csv. Default: stderr.
         -G | --nograph         : disable graphs on HTML output. Enabled by default.

diff --git a/doc/pgBadger.pod b/doc/pgBadger.pod
index cfb3dd9d7a1168ffd66d0ba8106a09b7e0dc4e64..30d42475c7d6fc096d16316aab522635a1e58cf8 100644 (file)
--- a/doc/pgBadger.pod
+++ b/doc/pgBadger.pod
@@ -25,6 +25,8 @@ Options:
     -c | --dbclient host   : only report on entries for the given client host.
     -C | --nocomment       : remove comments like /* ... */ from queries.
     -d | --dbname database : only report on entries for the given database.
+    -D | --dns-resolv      : client ip adresses are replaced by their DNS name.
+                             Be warned that this can really slow down pgBadger.
     -e | --end datetime    : end date/time for the data to be parsed in log.
     -f | --format logtype  : possible values: syslog,stderr,csv. Default: stderr.
     -G | --nograph         : disable graphs on HTML output. Enabled by default.

diff --git a/pgbadger b/pgbadger
index d165c4ba1482fe6f55e569ab21db59ef8f96993d..2d7dfb216319a8429728aa289b588ff1018ee288 100755 (executable)
--- a/pgbadger
+++ b/pgbadger
@@ -43,6 +43,7 @@ use File::Temp qw/ tempfile /;
 use IO::Handle;
 use IO::Pipe;
 use FileHandle;
+use Socket;
 
 $VERSION = '6.1';
 
@@ -60,6 +61,8 @@ my $graphid      = 1;
 my $NODATA       = '<div class="flotr-graph"><blockquote><b>NO DATASET</b></blockquote></div>';
 my $MAX_QUERY_LENGTH = 20480;
 my $terminate = 0;
+my %CACHE_DNS = ();
+my $DNSLookupTimeout = 1; # (in seconds)
 
 my $pgbadger_logo =
         '<img src="data:image/png;base64,
@@ -223,6 +226,7 @@ my $anonymize               = 0;
 my $noclean                 = 0;
 my $retention               = 0;
 my $bar_graph               = 0;
+my $dns_resolv              = 0;
 
 my $NUMPROGRESS = 10000;
 my @DIMENSIONS  = (800, 300);
@@ -306,6 +310,7 @@ my $result = GetOptions(
        "c|dbclient=s"             => \@dbclient,
        "C|nocomment!"             => \$remove_comment,
        "d|dbname=s"               => \@dbname,
+       "D|dns-resolv!"            => \$dns_resolv,
        "e|end=s"                  => \$to,
        "f|format=s"               => \$format,
        "G|nograph!"               => \$nograph,
@@ -1549,6 +1554,8 @@ Options:
     -c | --dbclient host   : only report on entries for the given client host.
     -C | --nocomment       : remove comments like /* ... */ from queries.
     -d | --dbname database : only report on entries for the given database.
+    -D | --dns-resolv      : client ip adresses are replaced by their DNS name.
+                             Be warned that this can really slow down pgBadger.
     -e | --end datetime    : end date/time for the data to be parsed in log.
     -f | --format logtype  : possible values: syslog,stderr,csv. Default: stderr.
     -G | --nograph         : disable graphs on HTML output. Enabled by default.
@@ -2003,6 +2010,7 @@ sub process_file
                                $prefix_vars{'t_appname'} = $row->[22] || '';
                                $prefix_vars{'t_client'}  = $row->[4] || '';
                                $prefix_vars{'t_client'}  =~ s/:.*//;
+                               $prefix_vars{'t_client'} = _gethostbyaddr($prefix_vars{'t_client'}) if ($dns_resolv);
                                $prefix_vars{'t_host'}    = 'csv';
                                $prefix_vars{'t_pid'}     = $row->[3];
                                $prefix_vars{'t_session_line'} = $row->[5];
@@ -2151,6 +2159,7 @@ sub process_file
                                        next if (!&check_incremental_position($prefix_vars{'t_timestamp'}, $line));
                                        $cur_pid = $prefix_vars{'t_pid'};
                                        $goon = 1;
+                                       $prefix_vars{'t_client'} = _gethostbyaddr($prefix_vars{'t_client'}) if ($dns_resolv);
 
                                        # Store the current timestamp of the log line
                                        &store_current_timestamp($prefix_vars{'t_timestamp'});
@@ -2283,6 +2292,7 @@ sub process_file
                                        # Jump to the last line parsed if required
                                        next if (!&check_incremental_position($prefix_vars{'t_timestamp'}, $line));
                                        $cur_pid = $prefix_vars{'t_pid'};
+                                       $prefix_vars{'t_client'} = _gethostbyaddr($prefix_vars{'t_client'}) if ($dns_resolv);
 
                                        # Store the current timestamp of the log line
                                        &store_current_timestamp($prefix_vars{'t_timestamp'});
@@ -8727,7 +8737,6 @@ sub validate_log_line
        if ($#dbclient >= 0) {
 
                # Log line does not match the required dbclient
-               $prefix_vars{'t_client'} ||= $prefix_vars{'t_hostport'};
                if (!$prefix_vars{'t_client'} || !grep(/^$prefix_vars{'t_client'}$/i, @dbclient)) {
                        delete $current_sessions{$prefix_vars{'t_pid'}};
                        return 0;
@@ -11126,7 +11135,31 @@ sub get_calendar
 
 }
 
+sub _gethostbyaddr
+{
+       my $ip = shift;
+
+       my $host = undef;
+       unless(exists $CACHE_DNS{$ip}) {
+               eval {
+                       local $SIG{ALRM} = sub { die "DNS lookup timeout.\n"; };
+                       alarm($DNSLookupTimeout);
+                       $host = gethostbyaddr(inet_aton($ip), AF_INET);
+                       alarm(0);
+               };
+               if ($@) {
+                       $CACHE_DNS{$ip} = undef;
+                       #printf "_gethostbyaddr timeout : %s\n", $ip;
+               }
+               else {
+                       $CACHE_DNS{$ip} = $host;
+                       #printf "_gethostbyaddr success : %s (%s)\n", $ip, $host;
+               }
+       }
+
+       return $CACHE_DNS{$ip} || $ip;
 
+}
 
 __DATA__