Skip to content

Instantly share code, notes, and snippets.

@jtai
Last active January 16, 2019 16:06
Show Gist options
  • Star 37 You must be signed in to star a gist
  • Fork 13 You must be signed in to fork a gist
  • Save jtai/1368338 to your computer and use it in GitHub Desktop.
Save jtai/1368338 to your computer and use it in GitHub Desktop.
Quick and dirty script to parse output of /usr/sbin/tcpdump -vvv -s 0 -l port 53
<?php
// quick and dirty argument parsing
foreach ($argv as $arg) {
if ($arg == '-f') {
define('FOLLOW', true);
}
if ($arg == '-h') {
define('HISTOGRAM', true);
}
}
if (!defined('FOLLOW')) {
define('FOLLOW', false);
}
if (!defined('HISTOGRAM')) {
define('HISTOGRAM', false);
}
$queries = array();
$output = array();
while (!feof(STDIN)) {
$line = fgets(STDIN, 8192);
if (!preg_match('/^\d{2}:\d{2}:\d{2}\.\d+ IP [^>]+ > .+/', $line)) {
$line .= fgets(STDIN, 8192);
}
// query
if (preg_match('/^(\d{2}:\d{2}:\d{2}\.\d+) IP [^>]+ > .+ (\d+)\+ (A(AAA)?)\? ([^ ]+)/', $line, $regs)) {
$qtimestamp = $regs[1];
$id = $regs[2];
$qtype = $regs[3];
$qsubject = rtrim($regs[5], '.');
$queries[$id] = array($qtimestamp, $qtype, $qsubject);
} else
// response
if (preg_match('/^(\d{2}:\d{2}:\d{2}\.\d+) IP [^>]+ > .+ (\d+).? q: (A(AAA)?)\? ([^ ]+)/', $line, $regs)) {
$rtimestamp = $regs[1];
$id = $regs[2];
$rtype = $regs[3];
$rsubject = rtrim($regs[5], '.');
if (isset($queries[$id])) {
list ($qtimestamp, $qtype, $qsubject) = $queries[$id];
if ($rtype == $qtype && $qsubject == $rsubject) {
$ms = (parse_timestamp($rtimestamp) - parse_timestamp($qtimestamp)) * 1000;
$bangs = max(0, ceil(log10($ms)));
output($qtimestamp, sprintf(
"%- 4s %- 50s % 8.03f ms%s",
$qtype, $qsubject, $ms, $bangs ? ' ' . str_repeat('!', $bangs) : ''
));
} else {
output($qtimestamp, sprintf(
"got response for query %s, but query type or subject don't match, query was %s %s, response was %s %s",
$id, $qtype, $qsubject, $rtype, $rsubject
));
}
} else {
output($rtimestamp, sprintf(
"got response to unsolicited %s query for %s (query %s)",
$rtype, $rsubject, $id
));
}
unset($queries[$id]);
}
}
fclose(STDIN);
if (!FOLLOW) {
// look for queries without responses
foreach ($queries as $id => $query) {
list ($qtimestamp, $qtype, $qsubject) = $query;
output($qtimestamp, sprintf(
"%- 4s %- 50s - ms **********",
$qtype, $qsubject
));
}
// collate buffered output
foreach ($output as $hm => $bucket) {
ksort($bucket);
foreach ($bucket as $sa => $data) {
list($timestamp, $message) = $data;
output_real($timestamp, $message);
}
}
// print histogram data
if (HISTOGRAM) {
echo "\n\n\n";
foreach ($output as $hm => $bucket) {
// count failed queries
$failed = 0;
foreach ($bucket as $sa => $data) {
list($timestamp, $message) = $data;
if (strpos($message, '- ms **********') !== false) {
$failed++;
}
}
printf("%s,%d\n", $hm, $failed);
}
}
}
function output($timestamp, $message)
{
if (FOLLOW) {
output_real($timestamp, $message);
flush();
} else {
// buffer output
$hm = substr($timestamp, 0, 5);
$sa = substr($timestamp, 6);
global $output;
$output[$hm][$sa] = array($timestamp, $message);
}
}
function output_real($timestamp, $message)
{
printf("%s %s\n", $timestamp, $message);
}
function parse_timestamp($timestamp)
{
$hours = substr($timestamp, 0, 2);
$minutes = substr($timestamp, 3, 2);
$seconds = substr($timestamp, 6, 2);
$after = substr($timestamp, 9);
return $hours * 3600 + $minutes * 60 + $seconds + (float) ('0.'.$after);
}
@gerr1t
Copy link

gerr1t commented Dec 2, 2015

Hi!

Simple question:

A response time like 160.063 ms, how do I interpret it?

  • 160.063 ms? (0.160063 Seconds)
  • 160063 ms? (160 Seconds)

Thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment