2 # $Cambridge: exim/src/src/eximstats.src,v 1.4 2004/12/21 11:33:55 steve Exp $
4 # Copyright (c) 2001 University of Cambridge.
5 # See the file NOTICE for conditions of use and distribution.
7 # Perl script to generate statistics from one or more Exim log files.
9 # Usage: eximstats [<options>] <log file> <log file> ...
11 # 1996-05-21: Ignore lines not starting with valid date/time, just in case
12 # these get into a log file.
13 # 1996-11-19: Add the -h option to control the size of the histogram,
14 # and optionally turn it off.
15 # Use some Perl 5 things; it should be everywhere by now.
16 # Add the Perl -w option and rewrite so no warnings are given.
17 # Add the -t option to control the length of the "top" listing.
18 # Add the -ne, -nt options to turn off errors and transport
20 # Add information about length of time on queue, and -q<list> to
21 # control the intervals and turn it off.
22 # Add count and percentage of delayed messages to the Received
24 # Show total number of errors.
25 # Add count and percentage of messages with errors to Received
27 # Add information about relaying and -nr to suppress it.
28 # 1997-02-03 Merged in some of the things Nigel Metheringham had done:
30 # Added received histogram as well as delivered
31 # Added local senders' league table
32 # Added local recipients' league table
33 # 1997-03-10 Fixed typo "destinationss"
34 # Allow for intermediate address between final and original
35 # when testing for relaying
36 # Give better message when no input
37 # 1997-04-24 Fixed bug in layout of error listing that was depending on
38 # text length (output line got repeated).
39 # 1997-05-06 Bug in option decoding when only one option.
40 # Overflow bug when handling very large volumes.
41 # 1997-10-28 Updated to handle revised log format that might show
42 # HELO name as well as host name before IP number
43 # 1998-01-26 Bugs in the function for calculating the number of seconds
44 # since 1970 from a log date
45 # 1998-02-02 Delivery to :blackhole: doesn't have a T= entry in the log
46 # line; cope with this, thereby avoiding undefined problems
47 # Very short log line gave substring error
48 # 1998-02-03 A routed delivery to a local transport may not have <> in the
49 # log line; terminate the address at white space, not <
50 # 1998-09-07 If first line of input was a => line, $thissize was undefined;
52 # 1998-12-21 Adding of $thissize from => line should have been adding $size.
53 # Oops. Should have looked more closely when fixing the previous
55 # 1999-11-12 Increased the field widths for printed integers; numbers are
56 # bigger than originally envisaged.
57 # 2001-03-21 Converted seconds() routine to use Time::Local, fixing a bug
58 # whereby seconds($timestamp) - id_seconds($id) gave an
60 # Added POD documentation.
61 # Moved usage instructions into help() subroutine.
62 # Added 'use strict' and declared all global variables.
63 # Added '-html' flag and resultant code.
64 # Added '-cache' flag and resultant code.
65 # Added add_volume() routine and converted all volume variables
66 # to use it, fixing the overflow problems for individual hosts
68 # Converted all volume output to GB/MB/KB as appropriate.
69 # Don't store local user stats if -nfl is specified.
70 # Modifications done by: Steve Campbell (<steve@computurn.com>)
71 # 2001-04-02 Added the -t_remote_users flag. Steve Campbell.
72 # 2001-10-15 Added the -domain flag. Steve Campbell.
73 # 2001-10-16 Accept files on STDIN or on the command line. Steve Campbell.
74 # 2001-10-21 Removed -domain flag and added -bydomain, -byhost, and -byemail.
75 # We now generate our main parsing subroutine as an eval statement
76 # which improves performance dramatically when not all the results
77 # are required. We also cache the last timestamp to time convertion.
79 # NOTE: 'Top 50 destinations by (message count|volume)' lines are
80 # now 'Top N (host|email|domain) destinations by (message count|volume)'
81 # where N is the topcount. Steve Campbell.
83 # 2001-10-30 V1.16 Joachim Wieland.
84 # Fixed minor bugs in add_volume() when taking over this version
85 # for use in Exim 4: -w gave uninitialized value warnings in
86 # two situations: for the first addition to a counter, and if
87 # there were never any gigabytes, thereby leaving the $gigs
89 # Initialized $last_timestamp to stop a -w uninitialized warning.
90 # Minor layout tweak for grand totals (nitpicking).
91 # Put the IP addresses for relaying stats in [] and separated by
92 # a space from the domain name.
93 # Removed the IPv4-specific address test when picking out addresses
94 # for relaying. Anything inside [] is OK.
96 # 2002-07-02 Philip Hazel
97 # Fixed "uninitialized variable" message that occurred for relay
98 # messages that arrived from H=[1.2.3.4] hosts (no name shown).
99 # This bug didn't affect the output.
101 # 2002-04-15 V1.17 Joachim Wieland.
102 # Added -charts, -chartdir. -chartrel options which use
103 # GD::Graph modules to create graphical charts of the statistics.
105 # 2002-04-15 V1.18 Steve Campbell.
106 # Added a check for $domain to to stop a -w uninitialized warning.
107 # Added -byemaildomain option.
108 # Only print HTML header links to included tables!
110 # 2002-08-02 V1.19 Steve Campbell.
111 # Changed the debug mode to dump the parser onto STDERR rather
112 # than STDOUT. Documented the -d flag into the help().
113 # Rejoined the divergent 2002-04-15 and 2002-07-02 releases.
115 # 2002-08-21 V1.20 Steve Campbell.
116 # Added the '-merge' option to allow merging of previous reports.
117 # Fixed a missing semicolon when doing -bydomain.
118 # Make volume charts plot the data gigs and bytes rather than just bytes.
119 # Only process log lines with $flag =~ /<=|=>|->|==|\*\*|Co/
120 # Converted Emaildomain to Edomain - the column header was too wide!
121 # This changes the text output slightly. You can revert to the old
122 # column widths by changing $COLUMN_WIDTHS to 7;
124 # 2002-09-04 V1.21 Andreas J Mueller
125 # Local deliveries domain now defaults to 'localdomain'.
126 # Don't match F=<From> when looking for the user.
128 # 2002-09-05 V1.22 Steve Campbell
129 # Fixed a perl 5.005 incompatibility problem ('our' variables).
131 # 2002-09-11 V1.23 Steve Campbell
132 # Stopped -charts option from throwing errors on null data.
133 # Don't print out 'Errors encountered' unless there are any.
135 # 2002-10-21 V1.23a Philip Hazel - patch from Tony Finch put in until
136 # Steve's eximstats catches up.
137 # Handle log files that include the timezone after the timestamp.
138 # Switch to assuming that log timestamps are in local time, with
139 # an option for UTC timestamps, as in Exim itself.
141 # 2003-02-05 V1.24 Steve Campbell
142 # Added in Sergey Sholokh's code to convert '<' and '>' characters
143 # in HTML output. Also added code to convert them back with -merge.
144 # Fixed timestamp offsets to convert to seconds rather than minutes.
145 # Updated -merge to work with output files using timezones.
146 # Added cacheing to speed up the calculation of timezone offsets.
148 # 2003-02-07 V1.25 Steve Campbell
149 # Optimised the usage of mktime() in the seconds subroutine.
150 # Removed the now redundant '-cache' option.
151 # html2txt() now explicitly matches HTML tags.
152 # Implemented a new sorting algorithm - the top_n_sort() routine.
153 # Added Danny Carroll's '-nvr' flag and code.
155 # 2003-03-13 V1.26 Steve Campbell
156 # Implemented HTML compliance changes recommended by Bernard Massot.
157 # Bug fix to allow top_n_sort() to handle null keys.
158 # Convert all domains and edomains to lowercase.
159 # Remove preceding dots from domains.
161 # 2003-03-13 V1.27 Steve Campbell
162 # Replaced border attributes with 'border=1', as recommended by
165 # 2003-06-03 V1.28 John Newman
166 # Added in the ability to skip over the parsing and evaulation of
167 # specific transports as passed to eximstats via the new "-nt/.../"
168 # command line argument. This new switch allows the viewing of
169 # not more accurate statistics but more applicable statistics when
170 # special transports are in use (ie; SpamAssassin). We need to be
171 # able to ignore transports such as this otherwise the resulting
172 # local deliveries are significantly skewed (doubled)...
174 # 2003-11-06 V1.29 Steve Campbell
175 # Added the '-pattern "Description" "/pattern/"' option.
177 # 2004-02-17 V1.30 Steve Campbell
178 # Added warnings if required GD::Graph modules are not available or
179 # insufficient -chart* options are specified.
181 # 2004-02-20 V1.31 Andrea Balzi
182 # Only show the Local Sender/Destination links if the tables exist.
184 # 2004-07-05 V1.32 Steve Campbell
185 # Fix '-merge -h0' divide by zero error.
187 # 2004-07-15 V1.33 Steve Campbell
188 # Documentation update - I've converted the subroutine
189 # documentation from POD to comments.
191 # 2004-12-10 V1.34 Steve Campbell
192 # Eximstats can now parse syslog lines as well as mainlog lines.
194 # 2004-12-20 V1.35 Wouter Verhelst
195 # Pie charts by volume were actually generated by count. Fixed.
200 eximstats - generates statistics from Exim mainlog or syslog files.
204 eximstats [Options] mainlog1 mainlog2 ... > report.txt
205 eximstats -html [Options] mainlog1 mainlog2 ... > report.html
206 eximstats -merge [Options] report.1.txt report.2.txt ... > weekly_report.txt
214 histogram divisions per hour. The default is 1, and
215 0 suppresses histograms. Valid values are:
217 0, 1, 2, 3, 5, 10, 15, 20, 30 or 60.
221 Don't display error information.
225 Don't display relaying information.
227 =item B<-nr>I</pattern/>
229 Don't display relaying information that matches.
233 Don't display transport information.
235 =item B<-nt>I</pattern/>
237 Don't display transport information that matches
241 List of times for queuing information single 0 item suppresses.
245 Display top <number> sources/destinations
246 default is 50, 0 suppresses top listing.
250 Omit local sources/destinations in top listing.
252 =item B<-t_remote_users>
254 Include remote users in the top source/destination listings.
258 Show results by sending host. This may be combined with
259 B<-bydomain> and/or B<-byemail> and/or B<-byedomain>. If none of these options
260 are specified, then B<-byhost> is assumed as a default.
264 Show results by sending domain.
265 May be combined with B<-byhost> and/or B<-byemail> and/or B<-byedomain>.
269 Show results by sender's email address.
270 May be combined with B<-byhost> and/or B<-bydomain> and/or B<-byedomain>.
272 =item B<-byemaildomain> or B<-byedomain>
274 Show results by sender's email domain.
275 May be combined with B<-byhost> and/or B<-bydomain> and/or B<-byemail>.
277 =item B<-pattern> I<Description> I</Pattern/>
279 Look for the specified pattern and count the number of lines in which it appears.
280 This option can be specified multiple times. Eg:
282 -pattern 'Refused connections' '/refused connection/'
287 This option allows eximstats to merge old eximstat reports together. Eg:
289 eximstats mainlog.sun > report.sun.txt
290 eximstats mainlog.mon > report.mon.txt
291 eximstats mainlog.tue > report.tue.txt
292 eximstats mainlog.wed > report.web.txt
293 eximstats mainlog.thu > report.thu.txt
294 eximstats mainlog.fri > report.fri.txt
295 eximstats mainlog.sat > report.sat.txt
296 eximstats -merge report.*.txt > weekly_report.txt
297 eximstats -merge -html report.*.txt > weekly_report.html
303 You can merge text or html reports and output the results as text or html.
307 You can use all the normal eximstat output options, but only data
308 included in the original reports can be shown!
312 When merging reports, some loss of accuracy may occur in the top I<n> lists.
313 This will be towards the ends of the lists.
317 The order of items in the top I<n> lists may vary when the data volumes
318 round to the same value.
324 Output the results in HTML.
328 Create graphical charts to be displayed in HTML output.
330 This requires the following modules which can be obtained
331 from http://www.cpan.org/modules/01modules.index.html
343 To install these, download and unpack them, then use the normal perl installation procedure:
350 =item B<-chartdir>I <dir>
352 Create the charts in the directory <dir>
354 =item B<-chartrel>I <dir>
356 Specify the relative directory for the "img src=" tags from where to include
361 Debug flag. This outputs the eval()'d parser onto STDOUT which makes it
362 easier to trap errors in the eval section. Remember to add 1 to the line numbers to allow for the
369 Eximstats parses exim mainlog and syslog files to output a statistical
370 analysis of the messages processed. By default, a text
371 analysis is generated, but you can request an html analysis
372 by using the B<-html> flag. See the help (B<-help>) to learn
373 about how to create charts from the tables.
377 There is a web site at http://www.exim.org - this contains details of the
378 mailing list exim-users@exim.org.
382 This program does not perfectly handle messages whose received
383 and delivered log lines are in different files, which can happen
384 when you have multiple mail servers and a message cannot be
385 immeadiately delivered. Fixing this could be tricky...
392 # use Time::Local; # PH/FANF
395 use vars qw($HAVE_GD_Graph_pie $HAVE_GD_Graph_linespoints);
396 eval { require GD::Graph::pie; };
397 $HAVE_GD_Graph_pie = $@ ? 0 : 1;
398 eval { require GD::Graph::linespoints; };
399 $HAVE_GD_Graph_linespoints = $@ ? 0 : 1;
402 ##################################################
404 ##################################################
405 # 'use vars' instead of 'our' as perl5.005 is still in use out there!
406 use vars qw(@tab62 @days_per_month $gig);
407 use vars qw($VERSION);
408 use vars qw($COLUMN_WIDTHS);
412 (0,1,2,3,4,5,6,7,8,9,0,0,0,0,0,0, # 0-9
413 0,10,11,12,13,14,15,16,17,18,19,20, # A-K
414 21,22,23,24,25,26,27,28,29,30,31,32, # L-W
415 33,34,35, 0, 0, 0, 0, 0, # X-Z
416 0,36,37,38,39,40,41,42,43,44,45,46, # a-k
417 47,48,49,50,51,52,53,54,55,56,57,58, # l-w
420 @days_per_month = (0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334);
421 $gig = 1024 * 1024 * 1024;
424 # How much space do we allow for the Hosts/Domains/Emails/Edomains column headers?
427 # Declare global variables.
428 use vars qw($total_received_data $total_received_data_gigs $total_received_count);
429 use vars qw($total_delivered_data $total_delivered_data_gigs $total_delivered_count);
430 use vars qw(%arrival_time %size %from_host %from_address);
431 use vars qw(%timestamp2time); #Hash of timestamp => time.
432 use vars qw($last_timestamp $last_time); #The last time convertion done.
433 use vars qw($last_date $date_seconds); #The last date convertion done.
434 use vars qw($last_offset $offset_seconds); #The last time offset convertion done.
435 use vars qw($localtime_offset);
436 use vars qw($i); #General loop counter.
437 use vars qw($debug); #Debug mode?
438 use vars qw($ntopchart); #How many entries should make it into the chart?
439 use vars qw($gddirectory); #Where to put files from GD::Graph
443 # The following are parameters whose values are
444 # set by command line switches:
445 use vars qw($show_errors $show_relay $show_transport $transport_pattern);
446 use vars qw($topcount $local_league_table $include_remote_users);
447 use vars qw($hist_opt $hist_interval $hist_number $volume_rounding);
448 use vars qw($relay_pattern @queue_times $html @user_patterns @user_descriptions);
450 use vars qw(%do_sender); #Do sender by Host, Domain, Email, and/or Edomain tables.
451 use vars qw($charts $chartrel $chartdir $charts_option_specified);
452 use vars qw($merge_reports); #Merge old reports ?
454 # The following are modified in the parse() routine, and
455 # referred to in the print_*() routines.
456 use vars qw($queue_more_than $delayed_count $relayed_unshown $begin $end);
457 use vars qw(%received_count %received_data %received_data_gigs);
458 use vars qw(%delivered_count %delivered_data %delivered_data_gigs);
459 use vars qw(%received_count_user %received_data_user %received_data_gigs_user);
460 use vars qw(%delivered_count_user %delivered_data_user %delivered_data_gigs_user);
461 use vars qw(%transported_count %transported_data %transported_data_gigs);
462 use vars qw(%remote_delivered %relayed %delayed %had_error %errors_count);
463 use vars qw(@queue_bin @remote_queue_bin @received_interval_count @delivered_interval_count);
464 use vars qw(@user_pattern_totals);
466 use vars qw(%report_totals);
471 ##################################################
473 ##################################################
476 #######################################################################
479 # $rounded_volume = volume_rounded($bytes,$gigabytes);
481 # Given a data size in bytes, round it to KB, MB, or GB
484 # Eg 12000 => 12KB, 15000000 => 14GB, etc.
486 # Note: I've experimented with Math::BigInt and it results in a 33%
487 # performance degredation as opposed to storing numbers split into
488 # bytes and gigabytes.
489 #######################################################################
501 if ($volume_rounding) {
505 $rounded = sprintf("%6d", $x);
507 elsif ($x < 10000000) {
508 $rounded = sprintf("%4dKB", ($x + 512)/1024);
511 $rounded = sprintf("%4dMB", ($x + 512*1024)/(1024*1024));
514 # Values between 1GB and 10GB are printed in MB
516 $rounded = sprintf("%4dMB", ($g * 1024) + ($x + 512*1024)/(1024*1024));
519 # Handle values over 10GB
520 $rounded = sprintf("%4dGB", $g + ($x + $gig/2)/$gig);
524 # We don't want any rounding to be done.
525 $rounded = sprintf("%4d", ($g * $gig) + $x);
532 #######################################################################
535 # un_round($rounded_volume,\$bytes,\$gigabytes);
537 # Given a volume in KB, MB or GB, as generated by volume_rounded(),
538 # do the reverse transformation and convert it back into Bytes and Gigabytes.
539 # These are added to the $bytes and $gigabytes parameters.
541 # Given a data size in bytes, round it to KB, MB, or GB
544 # EG: 500 => (500,0), 14GB => (0,14), etc.
545 #######################################################################
547 my($rounded,$bytes_sref,$gigabytes_sref) = @_;
549 if ($rounded =~ /(\d+)GB/) {
550 $$gigabytes_sref += $1;
552 elsif ($rounded =~ /(\d+)MB/) {
553 $$gigabytes_sref += $1 / 1024;
554 $$bytes_sref += (($1 % 1024 ) * 1024 * 1024);
556 elsif ($rounded =~ /(\d+)KB/) {
557 $$gigabytes_sref += $1 / (1024 * 1024);
558 $$bytes_sref += ($1 % (1024 * 1024) * 1024);
560 elsif ($rounded =~ /(\d+)/) {
561 $$gigabytes_sref += $1 / $gig;
562 $$bytes_sref += $1 % $gig;
565 #Now reduce the bytes down to less than 1GB.
566 add_volume($bytes_sref,$gigabytes_sref,0) if ($$bytes_sref > $gig);
570 #######################################################################
573 # add_volume(\$bytes,\$gigs,$size);
575 # Add $size to $bytes/$gigs where this is a number split into
576 # bytes ($bytes) and gigabytes ($gigs). This is significantly
577 # faster than using Math::BigInt.
578 #######################################################################
580 my($bytes_ref,$gigs_ref,$size) = @_;
581 $$bytes_ref = 0 if ! defined $$bytes_ref;
582 $$gigs_ref = 0 if ! defined $$gigs_ref;
583 $$bytes_ref += $size;
584 while ($$bytes_ref > $gig) {
591 #######################################################################
594 # $formatted_time = format_time($seconds);
596 # Given a time in seconds, break it down into
597 # weeks, days, hours, minutes, and seconds.
599 # Eg 12005 => 3h20m5s
600 #######################################################################
612 $p .= "$w"."w" if $w > 0;
613 $p .= "$d"."d" if $d > 0;
614 $p .= "$h"."h" if $h > 0;
615 $p .= "$m"."m" if $m > 0;
616 $p .= "$s"."s" if $s > 0 || $p eq "";
621 #######################################################################
624 # $seconds = unformat_time($formatted_time);
626 # Given a time in weeks, days, hours, minutes, or seconds, convert it to seconds.
628 # Eg 3h20m5s => 12005
629 #######################################################################
631 my($formated_time) = pop @_;
634 while ($formated_time =~ s/^(\d+)([wdhms]?)//) {
635 $time += $1 if ($2 eq '' || $2 eq 's');
636 $time += $1 * 60 if ($2 eq 'm');
637 $time += $1 * 60 * 60 if ($2 eq 'h');
638 $time += $1 * 60 * 60 * 24 if ($2 eq 'd');
639 $time += $1 * 60 * 60 * 24 * 7 if ($2 eq 'w');
645 #######################################################################
648 # $time = seconds($timestamp);
650 # Given a time-of-day timestamp, convert it into a time() value using
651 # POSIX::mktime. We expect the timestamp to be of the form
652 # "$year-$mon-$day $hour:$min:$sec", with month going from 1 to 12,
653 # and the year to be absolute (we do the necessary conversions). The
654 # timestamp may be followed with an offset from UTC like "+$hh$mm"; if the
655 # offset is not present, and we have not been told that the log is in UTC
656 # (with the -utc option), then we adjust the time by the current local
657 # time offset so that it can be compared with the time recorded in message
660 # To improve performance, we only use mktime on the date ($year-$mon-$day),
661 # and only calculate it if the date is different to the previous time we
662 # came here. We then add on seconds for the '$hour:$min:$sec'.
664 # We also store the results of the last conversion done, and only
665 # recalculate if the date is different.
667 # We used to have the '-cache' flag which would store the results of the
668 # mktime() call. However, the current way of just using mktime() on the
669 # date obsoletes this.
670 #######################################################################
674 # Is the timestamp the same as the last one?
675 return $last_time if ($last_timestamp eq $timestamp);
677 return 0 unless ($timestamp =~ /^((\d{4})\-(\d\d)-(\d\d))\s(\d\d):(\d\d):(\d\d)( ([+-])(\d\d)(\d\d))?/o);
679 unless ($last_date eq $1) {
681 my(@timestamp) = (0,0,0,$4,$3,$2);
682 $timestamp[5] -= 1900;
684 $date_seconds = mktime(@timestamp);
686 my $time = $date_seconds + ($5 * 3600) + ($6 * 60) + $7;
688 # SC. Use cacheing. Also note we want seconds not minutes.
689 #my($this_offset) = ($10 * 60 + $11) * ($9 . "1") if defined $8;
690 if (defined $8 && ($8 ne $last_offset)) {
692 $offset_seconds = ($10 * 60 + $11) * 60;
693 $offset_seconds = -$offset_seconds if ($9 eq '-');
698 #$time -= $this_offset;
699 $time -= $offset_seconds;
700 } elsif (defined $localtime_offset) {
701 $time -= $localtime_offset;
704 # Store the last timestamp received.
705 $last_timestamp = $timestamp;
712 #######################################################################
715 # $time = id_seconds($message_id);
717 # Given a message ID, convert it into a time() value.
718 #######################################################################
720 my($sub_id) = substr((pop @_), 0, 6);
722 my(@c) = split(//, $sub_id);
723 while($#c >= 0) { $s = $s * 62 + $tab62[ord(shift @c) - ord('0')] }
729 #######################################################################
730 # calculate_localtime_offset();
732 # $localtime_offset = calculate_localtime_offset();
734 # Calculate the the localtime offset from gmtime in seconds.
736 # $localtime = time() + $localtime_offset.
738 # These are the same semantics as ISO 8601 and RFC 2822 timezone offsets.
739 # (West is negative, East is positive.)
740 #######################################################################
742 # $localtime = gmtime() + $localtime_offset. OLD COMMENT
743 # This subroutine commented out as it's not currently in use.
745 #sub calculate_localtime_offset {
746 # # Pick an arbitrary date, convert it to localtime & gmtime, and return the difference.
747 # my (@sample_date) = (0,0,0,5,5,100);
748 # my $localtime = timelocal(@sample_date);
749 # my $gmtime = timegm(@sample_date);
750 # my $offset = $localtime - $gmtime;
754 sub calculate_localtime_offset {
755 # Assume that the offset at the moment is valid across the whole
756 # period covered by the logs that we're analysing. This may not
757 # be true around the time the clocks change in spring or autumn.
759 # mktime works on local time and gmtime works in UTC
760 my $local = mktime(gmtime($utc));
761 return $local - $utc;
765 #######################################################################
766 # print_queue_times();
768 # $time = print_queue_times($message_type,\@queue_times,$queue_more_than);
770 # Given the type of messages being output, the array of message queue times,
771 # and the number of messages which exceeded the queue times, print out
773 #######################################################################
774 sub print_queue_times {
776 my($string,$array,$queue_more_than) = @_;
781 my $cumulative_percent = 0;
782 #$queue_unknown += keys %arrival_time;
784 my $queue_total = $queue_more_than;
785 for ($i = 0; $i <= $#queue_times; $i++) { $queue_total += $$array[$i] }
787 my $temp = "Time spent on the queue: $string";
791 print "<hr><a name=\"$string time\"></a><h2>$temp</h2>\n";
792 print "<table border=0 width=\"100%\">\n";
794 print "<table border=1>\n";
795 print "<tr><th>Time</th><th>Messages</th><th>Percentage</th><th>Cumulative Percentage</th>\n";
796 $format = "<tr><td align=\"right\">%s %s</td><td align=\"right\">%d</td><td align=\"right\">%5.1f%%</td><td align=\"right\">%5.1f%%</td>\n";
800 printf("%s\n%s\n\n", $temp, "-" x length($temp));
801 $format = "%5s %4s %6d %5.1f%% %5.1f%%\n";
804 for ($i = 0; $i <= $#queue_times; $i++) {
807 my $percent = ($$array[$i] * 100)/$queue_total;
808 $cumulative_percent += $percent;
810 $printed_one? " " : "Under",
811 format_time($queue_times[$i]),
812 $$array[$i], $percent, $cumulative_percent);
813 if (!defined($queue_times[$i])) {
816 push(@chartdatanames,
817 ($printed_one? "" : "Under") . format_time($queue_times[$i]));
818 push(@chartdatavals, $$array[$i]);
823 if ($queue_more_than > 0) {
824 my $percent = ($queue_more_than * 100)/$queue_total;
825 $cumulative_percent += $percent;
828 format_time($queue_times[$#queue_times]),
829 $queue_more_than, $percent, $cumulative_percent);
831 push(@chartdatanames, "Over " . format_time($queue_times[$#queue_times]));
832 push(@chartdatavals, $queue_more_than);
834 #printf("Unknown %6d\n", $queue_unknown) if $queue_unknown > 0;
839 if ($HAVE_GD_Graph_pie && $charts) {
844 my $graph = GD::Graph::pie->new(200, 200);
847 if ($string =~ /all/) { $pngname = "queue_all.png"; $title = "Queue (all)"; }
848 if ($string =~ /remote/) { $pngname = "queue_rem.png"; $title = "Queue (remote)"; }
852 my $gd = $graph->plot(\@data) or warn($graph->error);
854 open(IMG, ">$chartdir/$pngname") or die $!;
858 print "<img src=\"$chartrel/$pngname\">";
861 print "</td></tr></table>\n";
868 #######################################################################
871 # print_histogram('Deliverieds|Messages received',@interval_count);
873 # Print a histogram of the messages delivered/received per time slot
875 #######################################################################
876 sub print_histogram {
878 my(@interval_count) = @_;
882 for ($i = 0; $i < $hist_number; $i++)
883 { $maxd = $interval_count[$i] if $interval_count[$i] > $maxd; }
885 my $scale = int(($maxd + 25)/50);
886 $scale = 1 if $scale == 0;
889 if ($text eq "Deliveries")
891 $type = ($scale == 1)? "delivery" : "deliveries";
895 $type = ($scale == 1)? "message" : "messages";
898 my($title) = sprintf("$text per %s (each dot is $scale $type)",
899 ($hist_interval == 60)? "hour" :
900 ($hist_interval == 1)? "minute" : "$hist_interval minutes");
903 print "<hr><a name=\"$text\"></a><h2>$title</h2>\n";
904 print "<table border=0 width=\"100%\">\n";
905 print "<tr><td><pre>\n";
908 printf("%s\n%s\n\n", $title, "-" x length($title));
913 for ($i = 0; $i < $hist_number; $i++)
915 my $c = $interval_count[$i];
917 # If the interval is an hour (the maximum) print the starting and
918 # ending hours as a label. Otherwise print the starting hour and
919 # minutes, which take up the same space.
924 $temp = sprintf("%02d-%02d", $hour, $hour + 1);
926 push(@chartdatanames, $temp);
932 { $temp = sprintf("%02d:%02d", $hour, $minutes) }
934 { $temp = sprintf(" :%02d", $minutes) }
936 push(@chartdatanames, $temp);
937 $minutes += $hist_interval;
944 push(@chartdatavals, $c);
945 printf(" %6d %s\n", $c, "." x ($c/$scale));
952 if ($HAVE_GD_Graph_linespoints && $charts) {
953 # calculate the graph
958 my $graph = GD::Graph::linespoints->new(300, 300);
963 x_labels_vertical => 1
966 if ($text =~ /Deliveries/) { $pngname = "histogram_del.png"; }
967 if ($text =~ /Messages/) { $pngname = "histogram_mes.png"; }
968 my $gd = $graph->plot(\@data) or warn($graph->error);
970 open(IMG, ">$chartdir/$pngname") or die $!;
974 print "<img src=\"$chartrel/$pngname\">";
977 print "</td></tr></table>\n";
983 #######################################################################
984 # print_league_table();
986 # print_league_table($league_table_type,\%message_count,\%message_data,\%message_data_gigs);
988 # Given hashes of message count and message data, which are keyed by
989 # the table type (eg by the sending host), print a league table
990 # showing the top $topcount (defaults to 50).
991 #######################################################################
992 sub print_league_table {
993 my($text,$m_count,$m_data,$m_data_gigs) = @_;
994 my($name) = ($topcount == 1)? "$text" : "$topcount ${text}s";
995 my($temp) = "Top $name by message count";
996 my(@chartdatanames) = ();
997 my(@chartdatavals) = ();
998 my $chartotherval = 0;
1002 print "<hr><a name=\"$text count\"></a><h2>$temp</h2>\n";
1003 print "<table border=0 width=\"100%\">\n";
1005 print "<table border=1>\n";
1006 print "<tr><th>Messages</th><th>Bytes</th><th>\u$text</th>\n";
1008 # Align non-local addresses to the right (so all the .com's line up).
1009 # Local addresses are aligned on the left as they are userids.
1010 my $align = ($text !~ /local/i) ? 'right' : 'left';
1011 $format = "<tr><td align=\"right\">%d</td><td align=\"right\">%s</td><td align=\"$align\" nowrap>%s</td>\n";
1014 printf("%s\n%s\n\n", $temp, "-" x length($temp));
1015 $format = "%7d %10s %s\n";
1019 foreach $key (top_n_sort($topcount,$m_count,$m_data_gigs,$m_data)) {
1022 $htmlkey =~ s/>/\>\;/g;
1023 $htmlkey =~ s/</\<\;/g;
1024 printf($format, $$m_count{$key}, volume_rounded($$m_data{$key},$$m_data_gigs{$key}), $htmlkey);
1027 printf($format, $$m_count{$key}, volume_rounded($$m_data{$key},$$m_data_gigs{$key}), $key);
1029 if (scalar @chartdatanames < $ntopchart)
1031 push(@chartdatanames, $key);
1032 push(@chartdatavals, $$m_count{$key});
1036 $chartotherval += $$m_count{$key};
1039 push(@chartdatanames, "Other");
1040 push(@chartdatavals, $chartotherval);
1045 print "</td><td>\n";
1046 if ($HAVE_GD_Graph_pie && $charts)
1048 # calculate the graph
1053 my $graph = GD::Graph::pie->new(300, 300);
1056 y_label => 'Amount',
1057 title => 'By count',
1059 my $gd = $graph->plot(\@data) or warn($graph->error);
1063 open(IMG, ">$chartdir/${temp}_count.png") or die $!;
1067 print "<img src=\"$chartrel/${temp}_count.png\">";
1070 print "</td><td>\n";
1071 print "</td></tr></table>\n";
1075 $temp = "Top $name by volume";
1077 print "<hr><a name=\"$text volume\"></a><h2>$temp</h2>\n";
1078 print "<table border=0 width=\"100%\">\n";
1080 print "<table border=1>\n";
1081 print "<tr><th>Messages</th><th>Bytes</th><th>\u$text</th>\n";
1084 printf("%s\n%s\n\n", $temp, "-" x length($temp));
1087 @chartdatanames = ();
1088 @chartdatavals = ();
1091 foreach $key (top_n_sort($topcount,$m_data_gigs,$m_data,$m_count)) {
1092 # The largest volume will be the first (top of the list).
1093 # If it has at least 1 gig, then just use gigabytes to avoid
1094 # risking an integer overflow when generating the pie charts.
1095 if ($$m_data_gigs{$key}) {
1100 $htmlkey =~ s/>/\>\;/g;
1101 $htmlkey =~ s/</\<\;/g;
1102 printf($format, $$m_count{$key}, volume_rounded($$m_data{$key},$$m_data_gigs{$key}), $htmlkey);
1105 printf($format, $$m_count{$key}, volume_rounded($$m_data{$key},$$m_data_gigs{$key}), $key);
1108 if (scalar @chartdatanames < $ntopchart) {
1110 if ($$m_data_gigs{$key}) {
1111 push(@chartdatanames, $key);
1112 push(@chartdatavals, $$m_data_gigs{$key});
1116 push(@chartdatanames, $key);
1117 push(@chartdatavals, $$m_data{$key});
1121 $chartotherval += ($use_gig) ? $$m_data_gigs{$key} : $$m_data{$key};
1124 push(@chartdatanames, "Other");
1125 push(@chartdatavals, $chartotherval);
1129 print "</td><td>\n";
1130 if ($HAVE_GD_Graph_pie && $charts) {
1131 # calculate the graph
1136 my $graph = GD::Graph::pie->new(300, 300);
1139 y_label => 'Volume' ,
1140 title => 'By Volume',
1142 my $gd = $graph->plot(\@data) or warn($graph->error);
1146 open(IMG, ">$chartdir/${temp}_volume.png") or die $!;
1150 print "<img src=\"$chartrel/${temp}_volume.png\">";
1153 print "</td><td>\n";
1154 print "</td></tr></table>\n";
1161 #######################################################################
1164 # @sorted_keys = top_n_sort($n,$href1,$href2,$href3);
1166 # Given a hash which has numerical values, return the sorted $n keys which
1167 # point to the top values. The second and third hashes are used as
1168 # tiebreakers. They all must have the same keys.
1170 # The idea behind this routine is that when you only want to see the
1171 # top n members of a set, rather than sorting the entire set and then
1172 # plucking off the top n, sort through the stack as you go, discarding
1173 # any member which is lower than your current n'th highest member.
1175 # This proves to be an order of magnitude faster for large hashes.
1176 # On 200,000 lines of mainlog it benchmarked 9 times faster.
1177 # On 700,000 lines of mainlog it benchmarked 13.8 times faster.
1179 # We assume the values are > 0.
1180 #######################################################################
1182 my($n,$href1,$href2,$href3) = @_;
1184 # PH's original sort was:
1186 # foreach $key (sort
1188 # $$m_count{$b} <=> $$m_count{$a} ||
1189 # $$m_data_gigs{$b} <=> $$m_data_gigs{$a} ||
1190 # $$m_data{$b} <=> $$m_data{$a} ||
1196 #We use a key of '_' to represent non-existant values, as null keys are valid.
1197 #'_' is not a valid domain, edomain, host, or email.
1198 my(@top_n_keys) = ('_') x $n;
1199 my($minimum_value1,$minimum_value2,$minimum_value3) = (0,0,0);
1201 my $n_minus_1 = $n - 1;
1202 my $n_minus_2 = $n - 2;
1204 # Pick out the top $n keys.
1205 my($key,$value1,$value2,$value3,$i,$comparison,$insert_position);
1206 while (($key,$value1) = each %$href1) {
1208 #print STDERR "key $key ($value1,",$href2->{$key},",",$href3->{$key},") <=> ($minimum_value1,$minimum_value2,$minimum_value3)\n";
1210 # Check to see that the new value is bigger than the lowest of the
1211 # top n keys that we're keeping.
1212 $comparison = $value1 <=> $minimum_value1 ||
1213 $href2->{$key} <=> $minimum_value2 ||
1214 $href3->{$key} <=> $minimum_value3 ||
1215 $top_n_key cmp $key;
1216 next unless ($comparison == 1);
1218 # As we will be using these values a few times, extract them into scalars.
1219 $value2 = $href2->{$key};
1220 $value3 = $href3->{$key};
1222 # This key is bigger than the bottom n key, so the lowest position we
1223 # will insert it into is $n minus 1 (the bottom of the list).
1224 $insert_position = $n_minus_1;
1226 # Now go through the list, stopping when we find a key that we're
1227 # bigger than, or we come to the penultimate position - we've
1228 # already tested bigger than the last.
1230 # Note: we go top down as the list starts off empty.
1231 # Note: stepping through the list in this way benchmarks nearly
1232 # three times faster than doing a sort() on the reduced list.
1233 # I assume this is because the list is already in order, and
1234 # we get a performance boost from not having to do hash lookups
1236 for ($i = 0; $i < $n_minus_1; $i++) {
1237 $top_n_key = $top_n_keys[$i];
1238 if ( ($top_n_key eq '_') ||
1239 ( ($value1 <=> $href1->{$top_n_key} ||
1240 $value2 <=> $href2->{$top_n_key} ||
1241 $value3 <=> $href3->{$top_n_key} ||
1242 $top_n_key cmp $key) == 1
1245 $insert_position = $i;
1250 # Remove the last element, then insert the new one.
1251 $#top_n_keys = $n_minus_2;
1252 splice(@top_n_keys,$insert_position,0,$key);
1254 # Extract our new minimum values.
1255 $top_n_key = $top_n_keys[$n_minus_1];
1256 if ($top_n_key ne '_') {
1257 $minimum_value1 = $href1->{$top_n_key};
1258 $minimum_value2 = $href2->{$top_n_key};
1259 $minimum_value3 = $href3->{$top_n_key};
1263 # Return the top n list, grepping out non-existant values, just in case
1264 # we didn't have that many values.
1265 return(grep(!/^_$/,@top_n_keys));
1269 #######################################################################
1272 # $header = html_header($title);
1274 # Print our HTML header and start the <body> block.
1275 #######################################################################
1278 my $text = << "EoText";
1279 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
1282 <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-15">
1283 <title>$title</title>
1285 <body bgcolor="white">
1293 #######################################################################
1298 # Display usage instructions and exit.
1299 #######################################################################
1303 eximstats Version $VERSION
1305 Usage: eximstats [Options] mainlog1 mainlog2 ... > report.txt
1306 eximstats -html [Options] mainlog1 mainlog2 ... > report.html
1307 eximstats -merge [Options] report.1.txt report.2.txt ... > weekly_rep.txt
1308 eximstats -merge -html [Options] report.1.html ... > weekly_rep.html
1310 Parses exim mainlog or syslog files and generates a statistical analysis
1311 of the messages processed. Valid options are:
1313 -h<number> histogram divisions per hour. The default is 1, and
1314 0 suppresses histograms. Other valid values are:
1315 2, 3, 5, 10, 15, 20, 30 or 60.
1316 -ne don't display error information
1317 -nr don't display relaying information
1318 -nr/pattern/ don't display relaying information that matches
1319 -nt don't display transport information
1320 -nt/pattern/ don't display transport information that matches
1321 -nvr don't do volume rounding. Display in bytes, not KB/MB/GB.
1322 -q<list> list of times for queuing information
1323 single 0 item suppresses
1324 -t<number> display top <number> sources/destinations
1325 default is 50, 0 suppresses top listing
1326 -tnl omit local sources/destinations in top listing
1327 -t_remote_users show top user sources/destinations from non-local domains
1329 -byhost show results by sending host (default unless bydomain or
1330 byemail is specified)
1331 -bydomain show results by sending domain.
1332 -byemail show results by sender's email address
1333 -byedomain show results by sender's email domain
1335 -pattern "Description" /pattern/
1336 Count lines matching specified patterns and show them in
1337 the results. It can be specified multiple times. Eg:
1338 -pattern 'Refused connections' '/refused connection/'
1340 -merge merge previously generated reports into a new report
1342 -html output the results in HTML
1343 -charts Create charts (this requires the GD::Graph modules)
1344 -chartdir <dir> Create the charts' png files in the directory <dir>
1345 -chartrel <dir> Specify the relative directory for the "img src=" tags
1346 from where to include the charts in the html file
1347 -chartdir and -chartrel default to '.'
1349 -d Debug mode - dump the eval'ed parser onto STDERR.
1358 #######################################################################
1359 # generate_parser();
1361 # $parser = generate_parser();
1363 # This subroutine generates the parsing routine which will be
1364 # used to parse the mainlog. We take the base operation, and remove bits not in use.
1365 # This improves performance depending on what bits you take out or add.
1367 # I've tested using study(), but this does not improve performance.
1369 # We store our parsing routing in a variable, and process it looking for #IFDEF (Expression)
1370 # or #IFNDEF (Expression) statements and corresponding #ENDIF (Expression) statements. If
1371 # the expression evaluates to true, then it is included/excluded accordingly.
1372 #######################################################################
1373 sub generate_parser {
1375 my($ip,$host,$email,$edomain,$domain,$thissize,$size,$old,$new);
1376 my($tod,$m_hour,$m_min,$id,$flag);
1379 # Convert syslog lines to mainlog format.
1381 next unless s/^.*? exim\\b.*?: //;
1384 next if length($_) < 38;
1385 next unless /^(\\d{4}\\-\\d\\d-\\d\\d\\s(\\d\\d):(\\d\\d):\\d\\d( [-+]\\d\\d\\d\\d)?)/o;
1387 ($tod,$m_hour,$m_min) = ($1,$2,$3);
1390 my($extra) = defined($4)? 6 : 0;
1391 $id = substr($_, 20 + $extra, 16);
1392 $flag = substr($_, 37 + $extra, 2);
1395 # Watch for user specified patterns.
1396 my $user_pattern_index = 0;
1397 foreach (@user_patterns) {
1398 $user_pattern_totals[$user_pattern_index] = 0;
1399 $parser .= " \$user_pattern_totals[$user_pattern_index]++ if $_;\n";
1400 $user_pattern_index++;
1404 next unless ($flag =~ /<=|=>|->|==|\\*\\*|Co/);
1406 #Strip away the timestamp, ID and flag (which could be "Com" for completed)
1407 #This speeds up the later pattern matches.
1408 # $_ = substr($_, 40);
1410 $_ = substr($_, 40 + $extra); # PH
1412 # JN - Skip over certain transports as specified via the "-nt/.../" command
1413 # line switch (where ... is a perl style regular expression). This is
1414 # required so that transports that skew stats such as SpamAssassin can be
1416 #IFDEF ($transport_pattern)
1417 if (/\\sT=(\\S+)/) {
1418 next if ($1 =~ /$transport_pattern/o) ;
1420 #ENDIF ($transport_pattern)
1423 $host = "local"; #Host is local unless otherwise specified.
1424 $domain = "localdomain"; #Domain is localdomain unless otherwise specified.
1427 # Do some pattern matches to get the host and IP address.
1428 # We expect lines to be of the form "H=[IpAddr]" or "H=Host [IpAddr]" or
1429 # "H=Host (UnverifiedHost) [IpAddr]" or "H=(UnverifiedHost) [IpAddr]".
1430 # We do 2 separate matches to keep the matches simple and fast.
1431 if (/\\sH=(\\S+)/) {
1434 ($ip) = /\\sH=.*?(\\s\\[[^]]+\\])/;
1435 # If there is only an IP address, it will be in $host and $ip will be
1436 # unset. That is OK, because we only use $ip in conjunction with $host
1437 # below. But make it empty to avoid warning messages.
1438 $ip = "" if !defined $ip;
1440 #IFDEF ($do_sender{Domain})
1441 if ($host !~ /^\\[/ && $host =~ /^(\\(?)[^\\.]+\\.([^\\.]+\\..*)/) {
1442 # Remove the host portion from the DNS name. We ensure that we end up with
1443 # at least xxx.yyy. $host can be "(x.y.z)" or "x.y.z".
1444 $domain = lc("$1.$2");
1445 $domain =~ s/^\\.//; #Remove preceding dot.
1447 #ENDIF ($do_sender{Domain})
1451 #IFDEF ($do_sender{Email})
1452 $email = (/^(\S+)/) ? $1 : "";
1453 #ENDIF ($do_sender{Email})
1455 #IFDEF ($do_sender{Edomain})
1456 $edomain = (/^\S*?\\@(\S+)/) ? lc($1) : "";
1457 #ENDIF ($do_sender{Edomain})
1459 if ($tod lt $begin) {
1462 elsif ($tod gt $end) {
1467 if ($flag eq "<=") {
1468 $thissize = (/\\sS=(\\d+)( |$)/) ? $1 : 0;
1469 $size{$id} = $thissize;
1471 #IFDEF ($show_relay)
1472 if ($host ne "local") {
1473 # Save incoming information in case it becomes interesting
1474 # later, when delivery lines are read.
1475 my($from) = /^(\\S+)/;
1476 $from_host{$id} = "$host$ip";
1477 $from_address{$id} = $from;
1479 #ENDIF ($show_relay)
1481 #IFDEF ($local_league_table || $include_remote_users)
1485 #IFDEF ($local_league_table && $include_remote_users)
1486 { #Store both local and remote users.
1487 #ENDIF ($local_league_table && $include_remote_users)
1489 #IFDEF ($local_league_table && ! $include_remote_users)
1490 if ($host eq "local") { #Store local users only.
1491 #ENDIF ($local_league_table && ! $include_remote_users)
1493 #IFDEF ($include_remote_users && ! $local_league_table)
1494 if ($host ne "local") { #Store remote users only.
1495 #ENDIF ($include_remote_users && ! $local_league_table)
1497 $received_count_user{$user}++;
1498 add_volume(\\$received_data_user{$user},\\$received_data_gigs_user{$user},$thissize);
1501 #ENDIF ($local_league_table || $include_remote_users)
1503 #IFDEF ($do_sender{Host})
1504 $received_count{Host}{$host}++;
1505 add_volume(\\$received_data{Host}{$host},\\$received_data_gigs{Host}{$host},$thissize);
1506 #ENDIF ($do_sender{Host})
1508 #IFDEF ($do_sender{Domain})
1510 $received_count{Domain}{$domain}++;
1511 add_volume(\\$received_data{Domain}{$domain},\\$received_data_gigs{Domain}{$domain},$thissize);
1513 #ENDIF ($do_sender{Domain})
1515 #IFDEF ($do_sender{Email})
1516 $received_count{Email}{$email}++;
1517 add_volume(\\$received_data{Email}{$email},\\$received_data_gigs{Email}{$email},$thissize);
1518 #ENDIF ($do_sender{Email})
1520 #IFDEF ($do_sender{Edomain})
1521 $received_count{Edomain}{$edomain}++;
1522 add_volume(\\$received_data{Edomain}{$edomain},\\$received_data_gigs{Edomain}{$edomain},$thissize);
1523 #ENDIF ($do_sender{Edomain})
1525 $total_received_count++;
1526 add_volume(\\$total_received_data,\\$total_received_data_gigs,$thissize);
1528 #IFDEF ($#queue_times >= 0)
1529 $arrival_time{$id} = $tod;
1530 #ENDIF ($#queue_times >= 0)
1532 #IFDEF ($hist_opt > 0)
1533 $received_interval_count[($m_hour*60 + $m_min)/$hist_interval]++;
1534 #ENDIF ($hist_opt > 0)
1537 elsif ($flag eq "=>") {
1538 $size = $size{$id} || 0;
1539 if ($host ne "local") {
1540 $remote_delivered{$id} = 1;
1543 #IFDEF ($show_relay)
1544 # Determine relaying address if either only one address listed,
1545 # or two the same. If they are different, it implies a forwarding
1546 # or aliasing, which is not relaying. Note that for multi-aliased
1547 # addresses, there may be a further address between the first
1550 if (defined $from_host{$id}) {
1551 if (/^(\\S+)(?:\\s+\\([^)]\\))?\\s+<([^>]+)>/) {
1552 ($old,$new) = ($1,$2);
1558 if ("\\L$new" eq "\\L$old") {
1559 ($old) = /^(\\S+)/ if $old eq "";
1560 my $key = "H=\\L$from_host{$id}\\E A=\\L$from_address{$id}\\E => " .
1561 "H=\\L$host\\E$ip A=\\L$old\\E";
1562 if (!defined $relay_pattern || $key !~ /$relay_pattern/o) {
1563 $relayed{$key} = 0 if !defined $relayed{$key};
1571 #ENDIF ($show_relay)
1575 #IFDEF ($local_league_table || $include_remote_users)
1576 #IFDEF ($local_league_table && $include_remote_users)
1577 { #Store both local and remote users.
1578 #ENDIF ($local_league_table && $include_remote_users)
1580 #IFDEF ($local_league_table && ! $include_remote_users)
1581 if ($host eq "local") { #Store local users only.
1582 #ENDIF ($local_league_table && ! $include_remote_users)
1584 #IFDEF ($include_remote_users && ! $local_league_table)
1585 if ($host ne "local") { #Store remote users only.
1586 #ENDIF ($include_remote_users && ! $local_league_table)
1588 if (my($user) = split((/\\s</)? " <" : " ", $_)) {
1589 if ($user =~ /^[\\/|]/) {
1590 my($parent) = $_ =~ /(<[^@]+@?[^>]*>)/;
1591 $user = "$user $parent" if defined $parent;
1593 $delivered_count_user{$user}++;
1594 add_volume(\\$delivered_data_user{$user},\\$delivered_data_gigs_user{$user},$size);
1597 #ENDIF ($local_league_table || $include_remote_users)
1599 #IFDEF ($do_sender{Host})
1600 $delivered_count{Host}{$host}++;
1601 add_volume(\\$delivered_data{Host}{$host},\\$delivered_data_gigs{Host}{$host},$size);
1602 #ENDIF ($do_sender{Host})
1603 #IFDEF ($do_sender{Domain})
1605 $delivered_count{Domain}{$domain}++;
1606 add_volume(\\$delivered_data{Domain}{$domain},\\$delivered_data_gigs{Domain}{$domain},$size);
1608 #ENDIF ($do_sender{Domain})
1609 #IFDEF ($do_sender{Email})
1610 $delivered_count{Email}{$email}++;
1611 add_volume(\\$delivered_data{Email}{$email},\\$delivered_data_gigs{Email}{$email},$size);
1612 #ENDIF ($do_sender{Email})
1613 #IFDEF ($do_sender{Edomain})
1614 $delivered_count{Edomain}{$edomain}++;
1615 add_volume(\\$delivered_data{Edomain}{$edomain},\\$delivered_data_gigs{Edomain}{$edomain},$size);
1616 #ENDIF ($do_sender{Edomain})
1618 $total_delivered_count++;
1619 add_volume(\\$total_delivered_data,\\$total_delivered_data_gigs,$size);
1621 #IFDEF ($show_transport)
1622 my $transport = (/\\sT=(\\S+)/) ? $1 : ":blackhole:";
1623 $transported_count{$transport}++;
1624 add_volume(\\$transported_data{$transport},\\$transported_data_gigs{$transport},$size);
1625 #ENDIF ($show_transport)
1627 #IFDEF ($hist_opt > 0)
1628 $delivered_interval_count[($m_hour*60 + $m_min)/$hist_interval]++;
1629 #ENDIF ($hist_opt > 0)
1633 elsif ($flag eq "==" && defined($size{$id}) && !defined($delayed{$id})) {
1638 elsif ($flag eq "**") {
1639 $had_error{$id} = 1 if defined ($size{$id});
1641 #IFDEF ($show_errors)
1642 $errors_count{$_}++;
1643 #ENDIF ($show_errors)
1647 elsif ($flag eq "Co") {
1649 #IFDEF ($#queue_times >= 0)
1650 #Note: id_seconds() benchmarks as 42% slower than seconds() and computing
1651 #the time accounts for a significant portion of the run time.
1653 if (defined $arrival_time{$id}) {
1654 $queued = seconds($tod) - seconds($arrival_time{$id});
1655 delete($arrival_time{$id});
1658 $queued = seconds($tod) - id_seconds($id);
1661 for ($i = 0; $i <= $#queue_times; $i++) {
1662 if ($queued < $queue_times[$i]) {
1664 $remote_queue_bin[$i]++ if $remote_delivered{$id};
1668 $queue_more_than++ if $i > $#queue_times;
1669 #ENDIF ($#queue_times >= 0)
1671 #IFDEF ($show_relay)
1672 delete($from_host{$id});
1673 delete($from_address{$id});
1674 #ENDIF ($show_relay)
1679 # We now do a 'C preprocessor style operation on our parser
1680 # to remove bits not in use.
1681 my(%defines_in_operation,$removing_lines,$processed_parser);
1682 foreach (split (/\n/,$parser)) {
1683 if ((/^\s*#\s*IFDEF\s*\((.*?)\)/i && ! eval $1) ||
1684 (/^\s*#\s*IFNDEF\s*\((.*?)\)/i && eval $1) ) {
1685 $defines_in_operation{$1} = 1;
1686 $removing_lines = 1;
1689 $processed_parser .= $_."\n" unless $removing_lines;
1691 if (/^\s*#\s*ENDIF\s*\((.*?)\)/i) {
1692 delete $defines_in_operation{$1};
1693 unless (keys %defines_in_operation) {
1694 $removing_lines = 0;
1698 print STDERR "# START OF PARSER:\n$processed_parser\n# END OF PARSER\n\n" if $debug;
1700 return $processed_parser;
1705 #######################################################################
1708 # parse($parser,\*FILEHANDLE);
1710 # This subroutine accepts a parser and a filehandle from main and parses each
1711 # line. We store the results into global variables.
1712 #######################################################################
1714 my($parser,$fh) = @_;
1716 if ($merge_reports) {
1717 parse_old_eximstat_reports($fh);
1728 #######################################################################
1733 # Print our headers and contents.
1734 #######################################################################
1737 my $title = "Exim statistics from $begin to $end";
1740 print html_header($title);
1742 print "<li><a href=\"#grandtotal\">Grand total summary</a>\n";
1743 print "<li><a href=\"#patterns\">User Specified Patterns</a>\n" if @user_patterns;
1744 print "<li><a href=\"#transport\">Deliveries by Transport</a>\n" if $show_transport;
1746 print "<li><a href=\"#Messages received\">Messages received per hour</a>\n";
1747 print "<li><a href=\"#Deliveries\">Deliveries per hour</a>\n";
1749 if ($#queue_times >= 0) {
1750 print "<li><a href=\"#all messages time\">Time spent on the queue: all messages</a>\n";
1751 print "<li><a href=\"#messages with at least one remote delivery time\">Time spent on the queue: messages with at least one remote delivery</a>\n";
1753 print "<li><a href=\"#Relayed messages\">Relayed messages</a>\n" if $show_relay;
1755 foreach ('Host','Domain','Email','Edomain') {
1756 next unless $do_sender{$_};
1757 print "<li><a href=\"#sending \l$_ count\">Top $topcount sending \l${_}s by message count</a>\n";
1758 print "<li><a href=\"#sending \l$_ volume\">Top $topcount sending \l${_}s by volume</a>\n";
1760 if ($local_league_table || $include_remote_users) {
1761 print "<li><a href=\"#local sender count\">Top $topcount local senders by message count</a>\n";
1762 print "<li><a href=\"#local sender volume\">Top $topcount local senders by volume</a>\n";
1764 foreach ('Host','Domain','Email','Edomain') {
1765 next unless $do_sender{$_};
1766 print "<li><a href=\"#\l$_ destination count\">Top $topcount \l$_ destinations by message count</a>\n";
1767 print "<li><a href=\"#\l$_ destination volume\">Top $topcount \l$_ destinations by volume</a>\n";
1769 if ($local_league_table || $include_remote_users) {
1770 print "<li><a href=\"#local destination count\">Top $topcount local destinations by message count</a>\n";
1771 print "<li><a href=\"#local destination volume\">Top $topcount local destinations by volume</a>\n";
1774 print "<li><a href=\"#errors\">List of errors</a>\n" if %errors_count;
1775 print "</ul>\n<hr>\n";
1784 #######################################################################
1785 # print_grandtotals();
1787 # print_grandtotals();
1789 # Print the grand totals.
1790 #######################################################################
1791 sub print_grandtotals {
1793 # Get the sender by headings and results. This is complicated as we can have
1794 # different numbers of columns.
1795 my($sender_txt_header,$sender_html_header,$sender_txt_format,$sender_html_format);
1796 my(@received_totals,@delivered_totals);
1797 foreach ('Host','Domain','Email','Edomain') {
1798 next unless $do_sender{$_};
1799 if ($merge_reports) {
1800 push(@received_totals, get_report_total($report_totals{Received},"${_}s"));
1801 push(@delivered_totals,get_report_total($report_totals{Delivered},"${_}s"));
1804 push(@received_totals,scalar(keys %{$received_data{$_}}));
1805 push(@delivered_totals,scalar(keys %{$delivered_data{$_}}));
1807 $sender_html_header .= "<th>${_}s</th>";
1808 $sender_txt_header .= " " x ($COLUMN_WIDTHS - length($_)) . $_ . 's';
1809 $sender_html_format .= "<td align=\"right\">%d</td>";
1810 $sender_txt_format .= " " x ($COLUMN_WIDTHS - 5) . "%6d";
1813 my($format1,$format2);
1816 <a name="grandtotal"></a>
1817 <h2>Grand total summary</h2>
1819 <tr><th>TOTAL</th><th>Volume</th><th>Messages</th>$sender_html_header<th colspan=2>At least one addr<br>Delayed</th><th colspan=2>At least one addr<br>Failed</th>
1822 $format1 = "<tr><td>%s</td><td align=\"right\">%s</td>$sender_html_format<td align=\"right\">%d</td>";
1823 $format2 = "<td align=\"right\">%d</td><td align=\"right\">%4.1f%%</td><td align=\"right\">%d</td><td align=\"right\">%4.1f%%</td>";
1826 my $sender_spaces = " " x length($sender_txt_header);
1831 $sender_spaces At least one address
1832 TOTAL Volume Messages $sender_txt_header Delayed Failed
1834 $format1 = " %-16s %9s %6d $sender_txt_format";
1835 $format2 = " %6d %4.1f%% %6d %4.1f%%",
1838 my($volume,$failed_count);
1839 if ($merge_reports) {
1840 $volume = volume_rounded($report_totals{Received}{Volume}, $report_totals{Received}{'Volume-gigs'});
1841 $total_received_count = get_report_total($report_totals{Received},'Messages');
1842 $failed_count = get_report_total($report_totals{Received},'Failed');
1843 $delayed_count = get_report_total($report_totals{Received},'Delayed');
1846 $volume = volume_rounded($total_received_data, $total_received_data_gigs);
1847 $failed_count = keys %had_error;
1852 printf("$format1$format2\n",'Received',$volume,$total_received_count,
1853 @received_totals,$delayed_count,
1854 ($total_received_count) ? ($delayed_count*100/$total_received_count) : 0,
1856 ($total_received_count) ? ($failed_count*100/$total_received_count) : 0);
1859 if ($merge_reports) {
1860 $volume = volume_rounded($report_totals{Delivered}{Volume}, $report_totals{Delivered}{'Volume-gigs'});
1861 $total_delivered_count = get_report_total($report_totals{Delivered},'Messages');
1864 $volume = volume_rounded($total_delivered_data, $total_delivered_data_gigs);
1866 printf("$format1\n\n",'Delivered',$volume,$total_delivered_count,@delivered_totals);
1867 print "</table>\n" if $html;
1871 #######################################################################
1872 # print_user_patterns()
1874 # print_user_patterns();
1876 # Print the counts of user specified patterns.
1877 #######################################################################
1878 sub print_user_patterns {
1882 print "<hr><a name=\"patterns\"></a><h2>User Specified Patterns</h2>\n";
1883 print "<table border=0 width=\"100%\">\n";
1885 print "<table border=1>\n";
1886 print "<tr><th> </th><th>Total</th>\n";
1887 $format1 = "<tr><td>%s</td><td align=\"right\">%d</td>";
1890 print "User Specified Patterns\n";
1891 print "-----------------------";
1893 $format1 = " %-18s %6d";
1897 if ($merge_reports) {
1898 # We are getting our data from previous reports.
1899 foreach $key (@user_descriptions) {
1900 my $count = get_report_total($report_totals{patterns}{$key},'Total');
1901 printf("$format1\n",$key,$count);
1905 # We are getting our data from mainlog files.
1906 my $user_pattern_index = 0;
1907 foreach $key (@user_descriptions) {
1908 printf("$format1\n",$key,$user_pattern_totals[$user_pattern_index]);
1909 $user_pattern_index++;
1919 #######################################################################
1920 # print_transport();
1922 # print_transport();
1924 # Print totals by transport.
1925 #######################################################################
1926 sub print_transport {
1928 my(@chartdatanames);
1929 my(@chartdatavals_count);
1930 my(@chartdatavals_vol);
1931 no integer; #Lose this for charting the data.
1934 print "<hr><a name=\"transport\"></a><h2>Deliveries by Transport</h2>\n";
1935 print "<table border=0 width=\"100%\">\n";
1937 print "<table border=1>\n";
1938 print "<tr><th> </th><th>Volume</th><th>Messages</th>\n";
1939 $format1 = "<tr><td>%s</td><td align=\"right\">%s</td><td align=\"right\">%d</td>";
1942 print "Deliveries by transport\n";
1943 print "-----------------------";
1944 print "\n Volume Messages\n";
1945 $format1 = " %-18s %6s %6d";
1949 if ($merge_reports) {
1950 # We are getting our data from previous reports.
1951 foreach $key (sort keys %{$report_totals{transport}}) {
1952 my $count = get_report_total($report_totals{transport}{$key},'Messages');
1953 printf("$format1\n",$key,
1954 volume_rounded($report_totals{transport}{$key}{Volume},$report_totals{transport}{$key}{'Volume-gigs'}),
1956 push(@chartdatanames, $key);
1957 push(@chartdatavals_count, $count);
1958 push(@chartdatavals_vol, $report_totals{transport}{$key}{'Volume-gigs'}*$gig + $report_totals{transport}{$key}{Volume} );
1962 # We are getting our data from mainlog files.
1963 foreach $key (sort keys %transported_data) {
1964 printf("$format1\n",$key,
1965 volume_rounded($transported_data{$key},$transported_data_gigs{$key}),
1966 $transported_count{$key});
1967 push(@chartdatanames, $key);
1968 push(@chartdatavals_count, $transported_count{$key});
1969 push(@chartdatavals_vol, $transported_data_gigs{$key}*$gig + $transported_data{$key});
1974 print "</td><td>\n";
1975 if ($HAVE_GD_Graph_pie && $charts)
1977 # calculate the graph
1980 \@chartdatavals_count
1982 my $graph = GD::Graph::pie->new(200, 200);
1984 x_label => 'Transport',
1985 y_label => 'Messages',
1986 title => 'By count',
1988 my $gd = $graph->plot(\@data) or warn($graph->error);
1990 open(IMG, ">$chartdir/transports_count.png") or die $!;
1994 print "<img src=\"$chartrel/transports_count.png\">";
1997 print "</td><td>\n";
1999 if ($HAVE_GD_Graph_pie && $charts) {
2004 my $graph = GD::Graph::pie->new(200, 200);
2006 title => 'By volume',
2008 my $gd = $graph->plot(\@data) or warn($graph->error);
2010 open(IMG, ">$chartdir/transports_vol.png") or die $!;
2014 print "<img src=\"$chartrel/transports_vol.png\">";
2017 print "</td></tr></table>\n";
2024 #######################################################################
2029 # Print our totals by relay.
2030 #######################################################################
2032 my $temp = "Relayed messages";
2033 print "<hr><a name=\"$temp\"></a><h2>$temp</h2>\n" if $html;
2034 if (scalar(keys %relayed) > 0 || $relayed_unshown > 0) {
2040 print "<table border=1>\n";
2041 print "<tr><th>Count</th><th>From</th><th>To</th>\n";
2042 $format = "<tr><td align=\"right\">%d</td><td>%s</td><td>%s</td>\n";
2045 printf("%s\n%s\n\n", $temp, "-" x length($temp));
2046 $format = "%7d %s\n => %s\n";
2050 foreach $key (sort keys %relayed) {
2051 my $count = $relayed{$key};
2054 my($one,$two) = split(/=> /, $key);
2055 printf($format, $count, $one, $two);
2058 print "</table>\n<p>\n" if $html;
2059 print "${spacing}Total: $shown (plus $relayed_unshown unshown)\n";
2062 print "No relayed messages\n";
2063 print "-------------------\n" unless $html;
2070 #######################################################################
2075 # Print our errors. In HTML, we display them as a list rather than a table -
2076 # Netscape doesn't like large tables!
2077 #######################################################################
2079 my $total_errors = 0;
2081 if (scalar(keys %errors_count) != 0) {
2082 my $temp = "List of errors";
2085 print "<hr><a name=\"errors\"></a><h2>$temp</h2>\n";
2086 print "<ul><li><b>Count - Error</b>\n";
2087 $format = "<li>%d - %s\n";
2090 printf("%s\n%s\n\n", $temp, "-" x length($temp));
2094 foreach $key (sort keys %errors_count) {
2097 $text =~ s/\s\s+/ /g; #Convert multiple spaces to a single space.
2098 $total_errors += $errors_count{$key};
2101 #Translate HTML tag characters. Sergey Sholokh.
2102 $text =~ s/\</\<\;/g;
2103 $text =~ s/\>/\>\;/g;
2105 printf($format,$errors_count{$key},$text);
2108 printf("%5d ", $errors_count{$key});
2109 while (length($text) > 65) {
2110 my($first,$rest) = $text =~ /(.{50}\S*)\s+(.+)/;
2112 printf("%s\n ", $first);
2115 printf("%s\n\n", $text);
2118 print "</ul>\n<p>\n" if $html;
2120 $temp = "Errors encountered: $total_errors";
2122 print "-" x length($temp),"\n" unless $html;
2128 #######################################################################
2129 # parse_old_eximstat_reports();
2131 # parse_old_eximstat_reports($fh);
2133 # Parse old eximstat output so we can merge daily stats to weekly stats and weekly to monthly etc.
2135 # To test that the merging still works after changes, do something like the following.
2136 # All the diffs should produce no output.
2138 # options='-bydomain -byemail -byhost -byedomain'
2139 # options="$options -pattern 'Completed Messages' /Completed/"
2140 # options="$options -pattern 'Received Messages' /<=/"
2142 # ./eximstats $options mainlog > mainlog.txt
2143 # ./eximstats $options -merge mainlog.txt > mainlog.2.txt
2144 # diff mainlog.txt mainlog.2.txt
2146 # ./eximstats $options -html mainlog > mainlog.html
2147 # ./eximstats $options -merge -html mainlog.txt > mainlog.2.html
2148 # diff mainlog.html mainlog.2.html
2150 # ./eximstats $options -merge mainlog.html > mainlog.3.txt
2151 # diff mainlog.txt mainlog.3.txt
2153 # ./eximstats $options -merge -html mainlog.html > mainlog.3.html
2154 # diff mainlog.html mainlog.3.html
2156 # ./eximstats $options -nvr mainlog > mainlog.nvr.txt
2157 # ./eximstats $options -merge mainlog.nvr.txt > mainlog.4.txt
2158 # diff mainlog.txt mainlog.4.txt
2160 # # double_mainlog.txt should have twice the values that mainlog.txt has.
2161 # ./eximstats $options mainlog mainlog > double_mainlog.txt
2162 #######################################################################
2163 sub parse_old_eximstat_reports {
2166 my(%league_table_value_entered, %league_table_value_was_zero, %table_order);
2169 if (/Exim statistics from ([\d\-]+ [\d:]+(\s+[\+\-]\d+)?) to ([\d\-]+ [\d:]+(\s+[\+\-]\d+)?)/) {
2170 $begin = $1 if ($1 lt $begin);
2171 $end = $3 if ($3 gt $end);
2173 elsif (/Grand total summary/) {
2174 # Fill in $report_totals{Received|Delivered}{Volume|Messages|Hosts|Domains|...|Delayed|DelayedPercent|Failed|FailedPercent}
2177 $_ = html2txt($_); #Convert general HTML markup to text.
2178 s/At least one addr//g; #Another part of the HTML output we don't want.
2180 # TOTAL Volume Messages Hosts Domains Delayed Failed
2181 # Received 26MB 237 177 23 8 3.4% 28 11.8%
2182 # Delivered 13MB 233 99 88
2183 if (/TOTAL\s+(.*?)\s*$/) {
2184 @fields = split(/\s+/,$1);
2185 #Delayed and Failed have two columns each, so add the extra field names in.
2186 splice(@fields,-1,1,'DelayedPercent','Failed','FailedPercent');
2188 elsif (/(Received|Delivered)\s+(.*?)\s*$/) {
2189 print STDERR "Parsing $_" if $debug;
2190 add_to_totals($report_totals{$1},\@fields,$2);
2192 last if (/Delivered/); #Last line of this section.
2196 elsif (/User Specified Patterns/i) {
2197 #User Specified Patterns
2198 #-----------------------
2202 while (<$fh>) { last if (/Total/); } #Wait until we get the table headers.
2204 print STDERR "Parsing $_" if $debug;
2205 $_ = html2txt($_); #Convert general HTML markup to text.
2206 if (/^\s*(.*?)\s+(\d+)\s*$/) {
2207 $report_totals{patterns}{$1} = {} unless (defined $report_totals{patterns}{$1});
2208 add_to_totals($report_totals{patterns}{$1},['Total'],$2);
2210 last if (/^\s*$/); #Finished if we have a blank line.
2214 elsif (/Deliveries by transport/i) {
2215 #Deliveries by transport
2216 #-----------------------
2218 # :blackhole: 70KB 51
2219 # address_pipe 655KB 1
2222 while (<$fh>) { last if (/Volume/); } #Wait until we get the table headers.
2224 print STDERR "Parsing $_" if $debug;
2225 $_ = html2txt($_); #Convert general HTML markup to text.
2226 if (/(\S+)\s+(\d+\S*\s+\d+)/) {
2227 $report_totals{transport}{$1} = {} unless (defined $report_totals{transport}{$1});
2228 add_to_totals($report_totals{transport}{$1},['Volume','Messages'],$2);
2230 last if (/^\s*$/); #Finished if we have a blank line.
2233 elsif (/(Messages received|Deliveries) per/) {
2234 # Messages received per hour (each dot is 2 messages)
2235 #---------------------------------------------------
2237 #00-01 106 .....................................................
2238 #01-02 103 ...................................................
2240 # Set a pointer to the interval array so we can use the same code
2241 # block for both messages received and delivered.
2242 my $interval_aref = ($1 eq 'Deliveries') ? \@delivered_interval_count : \@received_interval_count;
2243 my $reached_table = 0;
2245 $reached_table = 1 if (/^00/);
2246 next unless $reached_table;
2247 print STDERR "Parsing $_" if $debug;
2248 if (/^(\d+):(\d+)\s+(\d+)/) { #hh:mm start time format ?
2249 $$interval_aref[($1*60 + $2)/$hist_interval] += $3 if $hist_opt;
2251 elsif (/^(\d+)-(\d+)\s+(\d+)/) { #hh-hh start-end time format ?
2252 $$interval_aref[($1*60)/$hist_interval] += $3 if $hist_opt;
2254 else { #Finished the table ?
2260 elsif (/Time spent on the queue: (all messages|messages with at least one remote delivery)/) {
2261 #Time spent on the queue: all messages
2262 #-------------------------------------
2264 #Under 1m 217 91.9% 91.9%
2270 # Set a pointer to the queue bin so we can use the same code
2271 # block for both all messages and remote deliveries.
2272 my $bin_aref = ($1 eq 'all messages') ? \@queue_bin : \@remote_queue_bin;
2273 my $reached_table = 0;
2275 $_ = html2txt($_); #Convert general HTML markup to text.
2276 $reached_table = 1 if (/^\s*Under/);
2277 next unless $reached_table;
2278 my $previous_seconds_on_queue = 0;
2279 if (/^\s*(Under|Over|)\s+(\d+[smhdw])\s+(\d+)/) {
2280 print STDERR "Parsing $_" if $debug;
2281 my($modifier,$formated_time,$count) = ($1,$2,$3);
2282 my $seconds = unformat_time($formated_time);
2283 my $time_on_queue = ($seconds + $previous_seconds_on_queue) / 2;
2284 $previous_seconds_on_queue = $seconds;
2285 $time_on_queue = $seconds * 2 if ($modifier eq 'Over');
2287 for ($i = 0; $i <= $#queue_times; $i++) {
2288 if ($time_on_queue < $queue_times[$i]) {
2289 $$bin_aref[$i] += $count;
2293 # There's only one counter for messages going over the queue
2294 # times so make sure we only count it once.
2295 $queue_more_than += $count if (($bin_aref == \@queue_bin) && ($i > $#queue_times));
2298 last; #Finished the table ?
2303 elsif (/Relayed messages/) {
2307 # 1 addr.domain.com [1.2.3.4] a.user@domain.com
2308 # => addr2.domain2.com [5.6.7.8] a2.user2@domain2.com
2310 #<tr><td align="right">1</td><td>addr.domain.com [1.2.3.4] a.user@domain.com </td><td>addr2.domain2.com [5.6.7.8] a2.user2@domain2.com</td>
2312 my $reached_table = 0;
2315 unless ($reached_table) {
2316 last if (/No relayed messages/);
2317 $reached_table = 1 if (/^\s*\d/ || />\d+</);
2318 next unless $reached_table;
2320 if (/>(\d+)<.td><td>(.*?) ?<.td><td>(.*?)</) {
2321 update_relayed($1,$2,$3);
2323 elsif (/^\s*(\d+)\s+(.*?)\s*$/) {
2324 ($count,$sender) = ($1,$2);
2326 elsif (/=>\s+(.*?)\s*$/) {
2327 update_relayed($count,$sender,$1);
2330 last; #Finished the table ?
2335 elsif (/Top (.*?) by (message count|volume)/) {
2336 #Top 50 sending hosts by message count
2337 #-------------------------------------
2340 my($category,$by_count_or_volume) = ($1,$2);
2342 #As we show 2 views of each table (by count and by volume),
2343 #most (but not all) entries will appear in both tables.
2344 #Set up a hash to record which entries we have already seen
2345 #and one to record which ones we are seeing for the first time.
2346 if ($by_count_or_volume =~ /count/) {
2347 undef %league_table_value_entered;
2348 undef %league_table_value_was_zero;
2352 #As this section processes multiple different table categories,
2353 #set up pointers to the hashes to be updated.
2354 my($count_href,$data_href,$data_gigs_href);
2355 if ($category =~ /local sender/) {
2356 $count_href = \%received_count_user;
2357 $data_href = \%received_data_user;
2358 $data_gigs_href = \%received_data_gigs_user;
2360 elsif ($category =~ /sending (\S+?)s?\b/) {
2361 #Top 50 sending (host|domain|email|edomain)s
2362 #Top sending (host|domain|email|edomain)
2363 $count_href = \%{$received_count{"\u$1"}};
2364 $data_href = \%{$received_data{"\u$1"}};
2365 $data_gigs_href = \%{$received_data_gigs{"\u$1"}};
2367 elsif ($category =~ /local destination/) {
2368 $count_href = \%delivered_count_user;
2369 $data_href = \%delivered_data_user;
2370 $data_gigs_href = \%delivered_data_gigs_user;
2372 elsif ($category =~ /(\S+) destination/) {
2373 #Top 50 (host|domain|email|edomain) destinations
2374 #Top (host|domain|email|edomain) destination
2375 $count_href = \%{$delivered_count{"\u$1"}};
2376 $data_href = \%{$delivered_data{"\u$1"}};
2377 $data_gigs_href = \%{$delivered_data_gigs{"\u$1"}};
2380 my $reached_table = 0;
2382 $_ = html2txt($_); #Convert general HTML markup to text.
2383 $reached_table = 1 if (/^\s*\d/);
2384 next unless $reached_table;
2385 if (/^\s*(\d+)\s+(\S+)\s*(.*?)\s*$/) {
2386 my($count,$rounded_volume,$entry) = ($1,$2,$3);
2387 #Note: $entry fields can be both null and can contain spaces.
2389 #Add the entry into the %table_order hash if it has a rounded volume (KB/MB/GB).
2390 push(@{$table_order{$rounded_volume}{$by_count_or_volume}},$entry) if ($rounded_volume =~ /\D/);
2392 unless ($league_table_value_entered{$entry}) {
2393 $league_table_value_entered{$entry} = 1;
2394 unless ($$count_href{$entry}) {
2395 $$count_href{$entry} = 0;
2396 $$data_href{$entry} = 0;
2397 $$data_gigs_href{$entry} = 0;
2398 $league_table_value_was_zero{$entry} = 1;
2401 $$count_href{$entry} += $count;
2402 #Add the rounded value to the data and data_gigs hashes.
2403 un_round($rounded_volume,\$$data_href{$entry},\$$data_gigs_href{$entry});
2404 print STDERR "$category by $by_count_or_volume: added $count,$rounded_volume to $entry\n" if $debug;
2407 else { #Finished the table ?
2408 if ($by_count_or_volume =~ /volume/) {
2409 #Add a few bytes to appropriate entries to preserve the order.
2411 my($rounded_volume);
2412 foreach $rounded_volume (keys %table_order) {
2413 #For each rounded volume, we want to create a list which has things
2414 #ordered from the volume table at the front, and additional things
2415 #from the count table ordered at the back.
2416 @{$table_order{$rounded_volume}{volume}} = () unless defined $table_order{$rounded_volume}{volume};
2417 @{$table_order{$rounded_volume}{'message count'}} = () unless defined $table_order{$rounded_volume}{'message count'};
2419 map {$mark{$_} = 1} @{$table_order{$rounded_volume}{volume}};
2420 @order = @{$table_order{$rounded_volume}{volume}};
2421 map {push(@order,$_)} grep(!$mark{$_},@{$table_order{$rounded_volume}{'message count'}});
2423 my $bonus_bytes = $#order;
2424 $bonus_bytes = 511 if ($bonus_bytes > 511); #Don't go over the half-K boundary!
2425 while (@order and ($bonus_bytes > 0)) {
2426 my $entry = shift(@order);
2427 if ($league_table_value_was_zero{$entry}) {
2428 $$data_href{$entry} += $bonus_bytes;
2429 print STDERR "$category by $by_count_or_volume: added $bonus_bytes bonus bytes to $entry\n" if $debug;
2440 elsif (/List of errors/) {
2444 # 1 07904931641@one2one.net R=external T=smtp: SMTP error
2445 # from remote mailer after RCPT TO:<07904931641@one2one.net>:
2446 # host mail.one2one.net [193.133.192.24]: 550 User unknown
2448 #<li>1 - ally.dufc@dunbar.org.uk R=external T=smtp: SMTP error from remote mailer after RCPT TO:<ally.dufc@dunbar.org.uk>: host mail.dunbar.org.uk [216.167.89.88]: 550 Unknown local part ally.dufc in <ally.dufc@dunbar.org.uk>
2451 my $reached_table = 0;
2452 my($count,$error,$blanks);
2454 $reached_table = 1 if (/^( *|<li>)(\d+)/);
2455 next unless $reached_table;
2457 s/^<li>(\d+) -/$1/; #Convert an HTML line to a text line.
2458 $_ = html2txt($_); #Convert general HTML markup to text.
2461 $error .= ' ' . $1; #Join a multiline error.
2463 elsif (/^\s*(\d+)\s+(.*)/) {
2465 #Finished with a previous multiline error so save it.
2466 $errors_count{$error} = 0 unless $errors_count{$error};
2467 $errors_count{$error} += $count;
2469 ($count,$error) = ($1,$2);
2471 elsif (/Errors encountered/) {
2473 #Finished the section, so save our stored last error.
2474 $errors_count{$error} = 0 unless $errors_count{$error};
2475 $errors_count{$error} += $count;
2487 #######################################################################
2490 # update_relayed($count,$sender,$recipient);
2492 # Adds an entry into the %relayed hash. Currently only used when
2494 #######################################################################
2495 sub update_relayed {
2496 my($count,$sender,$recipient) = @_;
2498 #When generating the key, put in the 'H=' and 'A=' which can be used
2500 my $key = "H=$sender => H=$recipient";
2501 $key =~ s/ ([^=\s]+\@\S+|<>)/ A=$1/g;
2502 if (!defined $relay_pattern || $key !~ /$relay_pattern/o) {
2503 $relayed{$key} = 0 if !defined $relayed{$key};
2504 $relayed{$key} += $count;
2507 $relayed_unshown += $count;
2512 #######################################################################
2515 # add_to_totals(\%totals,\@keys,$values);
2517 # Given a line of space seperated values, add them into the provided hash using @keys
2520 # If the value contains a '%', then the value is set rather than added. Otherwise, we
2521 # convert the value to bytes and gigs. The gigs get added to I<Key>-gigs.
2522 #######################################################################
2524 my($totals_href,$keys_aref,$values) = @_;
2525 my(@values) = split(/\s+/,$values);
2526 my(@keys) = @$keys_aref; #Make a copy as we destroy the one we use.
2528 foreach $value (@values) {
2529 my $key = shift(@keys) or next;
2530 if ($value =~ /%/) {
2531 $$totals_href{$key} = $value;
2534 $$totals_href{$key} = 0 unless ($$totals_href{$key});
2535 $$totals_href{"$key-gigs"} = 0 unless ($$totals_href{"$key-gigs"});
2536 un_round($value, \$$totals_href{$key}, \$$totals_href{"$key-gigs"});
2537 print STDERR "Added $value to $key - $$totals_href{$key} , " . $$totals_href{"$key-gigs"} . "GB.\n" if $debug;
2542 #######################################################################
2543 # get_report_total();
2545 # $total = get_report_total(\%hash,$key);
2547 # If %hash contains values split into Units and Gigs, we calculate and return
2549 # $hash{$key} + 1024*1024*1024 * $hash{"${key}-gigs"}
2550 #######################################################################
2551 sub get_report_total {
2553 my($hash_ref,$key) = @_;
2554 if ($$hash_ref{"${key}-gigs"}) {
2555 return $$hash_ref{$key} + $gig * $$hash_ref{"${key}-gigs"};
2557 return $$hash_ref{$key} || 0;
2560 #######################################################################
2563 # $text_line = html2txt($html_line);
2565 # Convert a line from html to text. Currently we just convert HTML tags to spaces
2566 # and convert >, <, and tags back.
2567 #######################################################################
2571 # Convert HTML tags to spacing. Note that the reports may contain <Userid> and
2572 # <Userid@Domain> words, so explicitly specify the HTML tags we will remove
2573 # (the ones used by this program). If someone is careless enough to have their
2574 # Userid the same as an HTML tag, there's not much we can do about it.
2575 s/<\/?(html|head|title|body|h\d|ul|li|a\s+|table|tr|td|th|pre|hr|p|br)\b.*?>/ /og;
2577 s/\<\;/\</og; #Convert '<' to '<'.
2578 s/\>\;/\>/og; #Convert '>' to '>'.
2579 s/\ \;/ /og; #Convert ' ' to ' '.
2583 #######################################################################
2586 # $arg = get_next_arg();
2588 # Because eximstats arguments are often passed as variables,
2589 # we can't rely on shell parsing to deal with quotes. This
2590 # subroutine returns $ARGV[1] and does a shift. If $ARGV[1]
2591 # starts with a quote (' or "), and doesn't end in one, then
2592 # we append the next argument to it and shift again. We repeat
2593 # until we've got all of the argument.
2595 # This isn't perfect as all white space gets reduced to one space,
2596 # but it's as good as we can get! If it's esential that spacing
2597 # be preserved precisely, then you get that by not using shell
2599 #######################################################################
2602 my $matched_pattern = 0;
2604 $arg .= ' ' if $arg;
2605 $arg .= $ARGV[1]; shift(@ARGV);
2606 if ($arg !~ /^['"]/) {
2607 $matched_pattern = 1;
2610 if ($arg =~ s/^(['"])(.*)\1$/$2/) {
2611 $matched_pattern = 1;
2615 die "Mismatched argument quotes - <$arg>.\n" unless $matched_pattern;
2621 ##################################################
2623 ##################################################
2626 $last_timestamp = '';
2630 $show_transport = 1;
2632 $local_league_table = 1;
2633 $include_remote_users = 0;
2635 $volume_rounding = 1;
2636 $localtime_offset = calculate_localtime_offset(); # PH/FANF
2639 $charts_option_specified = 0;
2643 @queue_times = (60, 5*60, 15*60, 30*60, 60*60, 3*60*60, 6*60*60,
2644 12*60*60, 24*60*60);
2647 $offset_seconds = 0;
2651 while (@ARGV > 0 && substr($ARGV[0], 0, 1) eq '-')
2653 if ($ARGV[0] =~ /^\-h(\d+)$/) { $hist_opt = $1 }
2654 elsif ($ARGV[0] =~ /^\-ne$/) { $show_errors = 0 }
2655 elsif ($ARGV[0] =~ /^\-nr(.?)(.*)\1$/)
2657 if ($1 eq "") { $show_relay = 0 } else { $relay_pattern = $2 }
2659 elsif ($ARGV[0] =~ /^\-q([,\d\+\-\*\/]+)$/)
2661 @queue_times = split(/,/, $1);
2663 foreach $q (@queue_times) { $q = eval($q) + 0 }
2664 @queue_times = sort { $a <=> $b } @queue_times;
2665 @queue_times = () if ($#queue_times == 0 && $queue_times[0] == 0);
2667 elsif ($ARGV[0] =~ /^-nt$/) { $show_transport = 0 }
2668 elsif ($ARGV[0] =~ /^\-nt(.?)(.*)\1$/)
2670 if ($1 eq "") { $show_transport = 0 } else { $transport_pattern = $2 }
2672 elsif ($ARGV[0] =~ /^-t(\d+)$/) { $topcount = $1 }
2673 elsif ($ARGV[0] =~ /^-tnl$/) { $local_league_table = 0 }
2674 elsif ($ARGV[0] =~ /^-html$/) { $html = 1 }
2675 elsif ($ARGV[0] =~ /^-merge$/) { $merge_reports = 1 }
2676 elsif ($ARGV[0] =~ /^-charts$/) {
2678 warn "WARNING: CPAN Module GD::Graph::pie not installed. Obtain from www.cpan.org\n" unless $HAVE_GD_Graph_pie;
2679 warn "WARNING: CPAN Module GD::Graph::linespoints not installed. Obtain from www.cpan.org\n" unless $HAVE_GD_Graph_linespoints;
2681 elsif ($ARGV[0] =~ /^-chartdir$/) { $chartdir = $ARGV[1]; shift; $charts_option_specified = 1; }
2682 elsif ($ARGV[0] =~ /^-chartrel$/) { $chartrel = $ARGV[1]; shift; $charts_option_specified = 1; }
2683 elsif ($ARGV[0] =~ /^-cache$/) { } #Not currently used.
2684 elsif ($ARGV[0] =~ /^-byhost$/) { $do_sender{Host} = 1 }
2685 elsif ($ARGV[0] =~ /^-bydomain$/) { $do_sender{Domain} = 1 }
2686 elsif ($ARGV[0] =~ /^-byemail$/) { $do_sender{Email} = 1 }
2687 elsif ($ARGV[0] =~ /^-byemaildomain$/) { $do_sender{Edomain} = 1 }
2688 elsif ($ARGV[0] =~ /^-byedomain$/) { $do_sender{Edomain} = 1 }
2689 elsif ($ARGV[0] =~ /^-nvr$/) { $volume_rounding = 0 }
2690 elsif ($ARGV[0] =~ /^-d$/) { $debug = 1 }
2691 elsif ($ARGV[0] =~ /^--?h(elp)?$/){ help() }
2692 elsif ($ARGV[0] =~ /^-t_remote_users$/) { $include_remote_users = 1 }
2693 elsif ($ARGV[0] =~ /^-pattern$/)
2695 push(@user_descriptions,get_next_arg());
2696 push(@user_patterns,get_next_arg());
2698 elsif ($ARGV[0] =~ /^-utc$/)
2700 # We don't need this value if the log is in UTC.
2701 $localtime_offset = undef;
2705 print STDERR "Eximstats: Unknown or malformed option $ARGV[0]\n";
2711 # Check that all the charts options are specified.
2712 warn "-charts option not specified. Use -help for help.\n" if ($charts_option_specified && ! $charts);
2714 # Default to display tables by sending Host.
2715 $do_sender{Host} = 1 unless ($do_sender{Domain} || $do_sender{Email} || $do_sender{Edomain});
2718 for (my $i = 0; $i <= $#queue_times; $i++) {
2720 $remote_queue_bin[$i] = 0;
2723 # Compute the number of slots for the histogram
2727 if ($hist_opt > 60 || 60 % $hist_opt != 0)
2729 print "Eximstats: -h must specify a factor of 60\n";
2732 $hist_interval = 60/$hist_opt; #Interval in minutes.
2733 $hist_number = (24*60)/$hist_interval; #Number of intervals per day.
2734 @received_interval_count = (0) x $hist_number;
2735 @delivered_interval_count = (0) x $hist_number;
2738 #$queue_unknown = 0;
2740 $total_received_data = 0;
2741 $total_received_data_gigs = 0;
2742 $total_received_count = 0;
2744 $total_delivered_data = 0;
2745 $total_delivered_data_gigs = 0;
2746 $total_delivered_count = 0;
2748 $queue_more_than = 0;
2750 $relayed_unshown = 0;
2751 $begin = "9999-99-99 99:99:99";
2752 $end = "0000-00-00 00:00:00";
2754 foreach $section ('Received','Delivered') {
2755 foreach $type ('Volume','Messages','Delayed','Failed','Hosts','Domains','Emails','Edomains') {
2756 $report_totals{$section}{$type} = 0;
2760 # Generate our parser.
2761 my $parser = generate_parser();
2766 # Scan the input files and collect the data
2767 foreach my $file (@ARGV) {
2768 if ($file =~ /\.gz/) {
2769 unless (open(FILE,"gunzip -c $file |")) {
2770 print STDERR "Failed to gunzip -c $file: $!";
2774 elsif ($file =~ /\.Z/) {
2775 unless (open(FILE,"uncompress -c $file |")) {
2776 print STDERR "Failed to uncompress -c $file: $!";
2781 unless (open(FILE,$file)) {
2782 print STDERR "Failed to read $file: $!";
2786 #Now parse the filehandle, updating the global variables.
2787 parse($parser,\*FILE);
2792 #No files provided. Parse STDIN, updating the global variables.
2793 parse($parser,\*STDIN);
2797 if ($begin eq "9999-99-99 99:99:99") {
2798 print "**** No valid log lines read\n";
2802 # Output our results.
2804 print_grandtotals();
2806 # Print counts of user specified patterns if required.
2807 print_user_patterns() if @user_patterns;
2809 # Print totals by transport if required.
2810 print_transport() if $show_transport;
2812 # Print the deliveries per interval as a histogram, unless configured not to.
2813 # First find the maximum in one interval and scale accordingly.
2814 if ($hist_opt > 0) {
2815 print_histogram("Messages received", @received_interval_count);
2816 print_histogram("Deliveries", @delivered_interval_count);
2819 # Print times on queue if required.
2820 if ($#queue_times >= 0) {
2821 print_queue_times("all messages", \@queue_bin,$queue_more_than);
2822 print_queue_times("messages with at least one remote delivery",\@remote_queue_bin,$queue_more_than);
2825 # Print relay information if required.
2826 print_relay() if $show_relay;
2828 # Print the league tables, if topcount isn't zero.
2829 if ($topcount > 0) {
2830 foreach ('Host','Domain','Email','Edomain') {
2831 next unless $do_sender{$_};
2832 print_league_table("sending \l$_", $received_count{$_}, $received_data{$_},$received_data_gigs{$_});
2835 print_league_table("local sender", \%received_count_user,
2836 \%received_data_user,\%received_data_gigs_user) if ($local_league_table || $include_remote_users);
2837 foreach ('Host','Domain','Email','Edomain') {
2838 next unless $do_sender{$_};
2839 print_league_table("\l$_ destination", $delivered_count{$_}, $delivered_data{$_},$delivered_data_gigs{$_});
2841 print_league_table("local destination", \%delivered_count_user,
2842 \%delivered_data_user,\%delivered_data_gigs_user) if ($local_league_table || $include_remote_users);
2845 # Print the error statistics if required.
2846 print_errors() if $show_errors;
2849 print "</body>\n</html>\n"