Home > Cloud, Microsoft > Extending Microsoft OMS to monitor Squid Proxy running in Linux with a plugin – part 1/3 #MSOMS

Extending Microsoft OMS to monitor Squid Proxy running in Linux with a plugin – part 1/3 #MSOMS

Since Microsoft released OMS, I have been an early adopter and evangelist for the solution. Not only it is simple to deploy but it gives you a full spectrum of many of the workloads you have either on-premises or in the cloud and it does not matter which cloud. Be it Azure, AWS, Google and many others.

So, as I was advising on OMS for a customer, I found that they were running Squid Proxy servers. The Squid proxy server is one of the most famous proxy servers in the world and it has been utilised for years in many organisations. For that reason I then I decided to look at how OMS could leverage the monitoring for Squid.


As you can see here: https://github.com/Microsoft/OMS-Agent-for-Linux/tree/master/installer/conf/omsagent.d there are already many plugins for OMS to  monitor Windows and many Linux OS as well, DNS, Network, SQL, MySQL, Postgree, VMware, MongoDB, Security, Audit, Change Tracking and so on.

But, there was no Squid plugin and that’s where I brought back my past years of experience as a developer and although that was a long, long time go, I was able to developer in ruby a Squid plugin for Microsoft OMS.

How I developed it?

PART 1 : LOG Files

  1. I started but investigating the squid log on /var/log/squid/access.log and then I research REGEX expressions to extract information out of it. Below is a extract of it

1479696836.902    134 TCP_MISS/301 488 open http://cnn.com/ – HIER_DIRECT/ –
1479696848.110    242 TCP_MISS/400 486 open http://www.sydney.com/ – HIER_DIRECT/ text/html
1479696860.004    407 TCP_MISS/301 636 open http://www.7news.com.au/ – HIER_DIRECT/ text/html

The initial difficult part for me was of to decouple the date/time to get it on a human readable format. So, after long hours of research and playing along, I decided for the following REGEX :

 REGEX =/(?<eventtime>(\d+))\.\d+\s+(?<duration>(\d+))\s+(?<sourceip>(\d+\.\d+\.\d+\.\d+))\s+(?<cache>(\w+))\/(?<status>(\d+))\s+(?<bytes>(\d+)\s+)(?<response>(\w+)\s+)(?<url>([^\s]+))\s+(?<user>(\w+|\-))\s+(?<method>(\S+.\S+))/
(If you have a better one, please feel free to shot me)


  1. I then wrote a squidparserlog.rb in ruby to parse the Squid access.log file and turn it into a OMS format
class SquidLogParserLib
require ‘date’
require ‘etc’
require_relative ‘oms_common’
require ‘fluent/parser’
    def initialize(error_handler)
@error_handler = error_handler
    REGEX =/(?<eventtime>(\d+))\.\d+\s+(?<duration>(\d+))\s+(?<sourceip>(\d+\.\d+\.\d+\.\d+))\s+(?<cache>(\w+))\/(?<status>(\d+))\s+(?<bytes>(\d+)\s+)(?<response>(\w+)\s+)(?<url>([^\s]+))\s+(?<user>(\w+|\-))\s+(?<method>(\S+.\S+))/
    def parse(line)
      data = {}
time = Time.now.to_f
REGEX.match(line) { |match|
data[‘Host’] = OMS::Common.get_hostname
          timestamp = Time.at( match[‘eventtime’].to_i() )
data[‘EventTime’] = OMS::Common.format_time(timestamp)
data[‘EventDate’] = timestamp.strftime( ‘%Y-%m-%d’ )
data[‘Duration’] = match[‘duration’].to_i()
data[‘SourceIP’] = match[‘sourceip’]
data[‘cache’] = match[‘cache’]
data[‘status’] = match[‘status’]
data[‘bytes’] = match[‘bytes’].to_i()
data[‘httpresponse’] = match[‘response’]
data[‘bytes’] = match[‘bytes’].to_i()
data[‘url’] = match[‘url’]
data[‘user’] = match[‘user’]
data[‘method’] = match[‘method’]}
rescue => e
@error_handler.logerror(“Unable to parse the line #{e}”)
      return time, data
end   #def
   end   #class
3. Finally, I wrote the squid.conf for OMS
# enhanced parse log with date format , which pass the path for the log to the SquidLogParser and tag it as oms.api.Squid. By doing this, you will end up with 11 custom fields in OMS for the LOG TYPE Squid_CL
type tail
format SquidLogParser
path /var/log/squid/access.log
pos_file /var/opt/microsoft/omsagent/state/var_log_squid_access.pos
tag oms.api.Squid
log_level error


On my next article I will go through the next part, which is getting Squid Proxy Statistics in OMS, along with the full code.



  1. November 24, 2016 at 10:17

    Reblogged this on Diário de Bordo.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: