Big companies usually have IT staff that is monitoring the server infrastructure 24/7, something that most website and server owners cannot (and do not have the desire to) afford. Some webhosting companies provide 24/7 monitoring of their dedicated servers, virtual servers and web hosting packages but the effectiveness varies a lot and the account owner is usually not informed at once if something goes down (and back up again).

A third alternative are so called web monitoring services. These services are web based and offer a set of tools to monitor web servers and websites. Most of these services offer email, sms and even phone notifications as soon as the server goes down.

Pingdom is one of those services. The web monitoring service offers various http, network and email checks that can be configured easily. The number of services that are monitored depends on the user’s account level. Pingdom offers a free account that is good for monitoring one web service on one domain. The basic and business accounts raise those limits to 5 respectively 30 checks with the option to buy extra checks if needed.

Website Monitoring Setup

Pingdom does not require any installations or configurations on the web server or website that the user wants to monitor. All monitoring services are setup on the Pingdom website.

It starts by selecting one of the available check types.

This ranges from http checks over network checks that test the DNS, TCP port or UDP to email checks that can check Pop3, SMTP and IMAP connections. The check resolution, that is the interval in which the check is performed can be selected in this step as well. The interval ranges from every minute to every 60 minutes.

The next options differ depending on the choice in the first step. A http check for instance requires only the url of the website that should be checked. Entering the url completes this step which does provide optional settings which can help a user in performing specific tasks.

The Check For String On Page parameter for instance can easily be used to check if a database is up and running. To do that the webmaster needs to write (or copy and paste) a small script that displays certain text on a custom url if the website is up or down. This can for instance be UP if it is up and DOWN if it is down. The string check can be configured to report downtime if the DOWN string appeared on the page of the website or server.

Other optional settings in this step include configuring a username and password for websites that require authentication and changing the user-agent of the Pingdom service (which might make sense if third-party websites are monitored).

The final step of the process configured the notification options. Pingdom uses a contact database that can be filled with contacts that should be notified if a check fails. This can always be the same user or individual webmasters or system administrators depending on the website or even check that is performed.

Pingdom supports four types of notifications: SMS, Email, Twitter and iPhone. The number of available SMS depends on the account type. The free account comes with 20 SMS with the option to buy additional SMS if needed. An option that is missing is phone support which is offered by some other website monitoring services.

Reporting

Pingdom starts monitoring the configured services as soon as they have been created by the user. Reporting is one of the strengths of Pingdom.

The overview displays the most important data at a glance including the uptime, downtime and response time. This report can be filtered by data to provide an overview of a specific time frame.

Detailed reports are available for each website that is listed in the overview and from the sidebar menu.

Pingdom operates servers in many countries in the world that are used to monitor the configured websites, servers and services. Successful and failed attempts to connect will be written to an extensive log file that provides error analysis for some failed attempts. This includes information like the date and time of the operation, the datacenter used to perform the check and a traceroute which can be very helpful for the error analysis.

Pingdom Review Verdict

Pingdom offers a sophisticated web monitoring service with excellent reporting. The service is aimed at webmasters, system administrators and professionals mainly which is reflected in the interface which could use additional help overlays to guide the user through the creation. It would also be nice if they would add an option to receive phone notifications and if they would provide basic scripts to test specific services like MySQL databases.

But those are marginal and the overall experience and service lives up to its promises. The free basic account should be enough for many users who only run one website. Professionals on the other hand can use the basic and business account along with the optional services to monitor all their servers and websites.

Christmas Giveaway

We are giving away 10 Pingdom Basic Accounts that are paid for for one year. Please let us know if you currently use a website monitoring service and if so which to enter the draw for the accounts.

Raju is giving away two programs today: Corel Digital Studio 2010 and Corel Painter Essentials 4.

Beecher is giving away a copy of Team Fortress 2, best online fps game since UT99.

In this article you will see how to install Nagios on a working Ubuntu Server installation. The release I used was 9.04, but you should be able to use 8.04 just as easily.

Features

You want features? Nagios is full of them. Nagios can monitor your entire network, you can handle problem remediation, you can plan network downtime, you can watch hosts, systems, services, applications..there’s very little you can’t do with Nagios.

But let’s think about this installation. The best (and really only good) way to install Nagios is from source. This means you are going to have to do some compliation. Don’t worry, it’s not hard…when you have a step-by-step guide that is.

Before you start

If you are using, as I did, a Ubuntu server installation, you will need to install a few tools first. So from the command line (on your Ubuntu server) issue the following commands:

sudo apt-get install php5-gd

sudo apt-get install gcc

sudo apt-get install make

The above commands will install the tools you need in order to get Nagios installed. If you don’t install the above, you won’t be able to complete the installation.

Installing Nagios – preflight

Before you actually run the installation you will need to take care of some user/group accounting first.

Create the user nagios:

sudo useradd -m nagios

Give the nagios user a password:

sudo passwd nagios

NOTE: You will have to enter the new password twice for the above command.

Create the group nagios:
sudo groupadd nagios

Add the user nagios to the group nagios:

sudo usermod -G nagios nagios

Create the group nagcmd:

sudo groupadd nagcmd

Add the user nagios to the group nagcmd:

sudo usermod -a -G nagcmd nagios

Add the Apache user to the group nagcmd:

sudo usermod -a -G nagcmd www-data

Installing Nagios and Nagios Plugins

The first thing to do is to download the files you need. Issue the commands:

wget http://osdn.dl.sourceforge.net/sourceforge/nagios/nagios-3.0.6.tar.gz

and

wget http://osdn.dl.sourceforge.net/sourceforge/nagiosplug/nagios-
plugins-1.4.11.tar.gz

to download the files you need.

Untar the Nagios file with:

tar xvzf nagios-3.0.6.tar.gz

Change into the newly created Nagios directory with the command:

cd nagios-3.0.6

Now run the configure script using the nagcmd defined as the command group with the command:

sudo ./configure –with-command-group=nagcmd

Time to compile:

sudo make all

Time to install everything:

make install

make install-init

make install-config

make install-commandmode

Almost ready to finalize the installation of Nagios (before moving on to installing the plugins.) There is one simple configuration to take care of (for the basic installation). Open up the /usr/local/nagios/etc/objects/contacts.cfg file and add your email address as the contact email address. That’s the only configuration to change for now.

Configuring for the Web interface

There are just a few commands to run to configure the Web interface:

make install-webconf

htpasswd -c /usr/local/nagios/etc/htpasswd.users nagiosadmin

/etc/init.d/apache2 reload

The final command above simply restarts Apache.

Install the plugins

Before you log into your Nagios installation you have to install your plugins first. Change to the directory you downloaded the Nagios files to and untar the plugins file:

tar xvzf nagios-plugins-1.4.11.tar.gz

Now change into the newly created directory:

cd nagios-plugins-1.4.11

Compile the plugins with the command:

sudo ./configure –with-nagios-user=nagios –with-nagios-group=nagios

Now install the plugins with the following commands:

make

make install

Fire it up!

Of course you want to make sure Nagios starts any time the system is restarted. Do this with the command:

sudo ln -s /etc/init.d/nagios /etc/rcS.d/S99nagios

And finally, before you start up Nagios, make sure there are no errors with the command:

sudo /usr/local/nagios/bin/nagios -v /usr/local/nagios/etc/nagios.cfg

If you see no errors reported, it’s time to start it up. Issue the command:

sudo /etc/init.d/nagios start

Nagios is ready to go. Point your browser to:

http://ADDRESS_OF_NAGIOS_SERVER/nagios

You should be ready to enjoy all that is Nagios. Congratulations.

The main interface of the monitoring software is divided into a quick server check option at the bottom which can be used to check if services such as HTTP, FTP or POP are running on a IP or domain. The networking monitoring software will automatically display the results in the same interface. Probably more interesting than the quick check function is the option to add servers permanently to the monitoring tool.

The user needs to enter a server name which is used for identification and a server address which is usually an IP address. A list of services are preselected and will be checked unless unchecked by the user. Alerts can be enabled or disabled in the same interface as well. These do have to be configured separately and can inform the user by email, screen message, executing a program or playing a sound.

There is also the possibility to add http url checks next to the default services that can be monitored by Open Monitor. This allows a user to specify an url, port, alerts and words that the page should include or exclude. This can be used for example to verify that a page is up by checking for a word on that page that is always appearing. Can also be used to run scripts that output a simple message in the end, e.g. a database check that is displaying ok if the database is up and running.

Open Monitor is a straightforward, easy to use, network monitoring software that can monitor popular services. Alternative programs are the server monitoring software IT Service View, Newt, the website monitoring software SiteMonitor or the Microsoft Network Monitor.

IT Service View is probably the server monitoring software with the broadest range of monitoring possibilities. They range from monitoring websites to Linux and Windows Server 2003 / 2008 servers to Microsoft Exchange Servers to Oracle servers and log files. The server monitoring is compatible with Microsoft Windows XP, Vista and Server 2003. It will display a wizard after installation that is walking the user through the configuration of the program.

The less complex monitoring services take only a few seconds to setup. Setting up a website monitoring service in the software that monitored the Ghacks database did not take longer than a minute. It has to be noted that the script that checks the database is already installed on the Ghacks server, the implementation itself was straightforward.

Several parameters can be configured in the setup phase. It is for example possible to define a maximum response time, time out, expected http codes, error texts and text that the website has to contain.

The even responses are highly configurable as well. They are divided into the possible scenarios “no response”, “error”, “error text”, “text not found”, “time out” and “slow web page” when configuring the website monitoring service.

Each of these can launch a series of responses including sending emails, running local programs, shutting down or rebooting windows, managing a Windows service and launching a Windows process.

Some or all of these event responses are displayed whenever configuring a new monitoring service in the server monitoring software.

The monitoring software displays various statistics about each configured job including graphs, log entries and script outputs.

The server monitoring services can monitor key variables on the servers including free computer memory, cpu utilization, service and process statuses and free disk space. Some of the services require that a script is running on the server so that the information can be analyzed. A Linux server has to run the net-snmp suite for example for It Service View to receive the data.

IT Service View is a sophisticated server monitoring software for IT professionals and small businesses who want to monitor their websites and servers.

Not everyone wants to pay to monitor websites though. The free portable software program SiteMonitor provides a local solution for the Windows operating system. It requires the Microsoft .net Framework 2.0 and can be configured to monitor selected websites and notify the user if they become unreachable. It is a good alternative as long as it is run on a computer system that is online 24/7.

New urls can be added easily to SiteMonitor. All that it takes is to add a descriptive name, an url and HTML to match which is simply content that should appear on the website. The website monitoring software will check every 30 minutes by default; To much for important websites. The interval can be changed in the options. The options are also the place where the notification preferences can be edited. The website monitor provides two different types of reactions to website connection problems:

• Run a script or executable, e.g. a script that will restart a server or service
• Send email to notify a user about the connection problem

The amount of failures before either one of the reactions will be initiated can be configured in the same menu as well.

Below is the small php script that we use to check if the database is up and running on Ghacks:

<?php
$dbhost = 'localhost';
$dbuser = 'username';
$dbpass = 'password';

$conn = mysql_connect($dbhost, $dbuser, $dbpass) or die ('Error');

$dbname = 'databasename';
mysql_select_db($dbname);

if ($conn) {echo "ok";}
else {echo "error";}

?>

replace username with the database username, password with the password of the username and databasename with the name of the database. Save it as database.php, upload it to your website and load the new file. It should display ok if the database is up and error if it is down. The url that points to the script can then be added to the website monitoring software to check if it returns ok. If it does the website is up, if it does not it is down.

Alternatives are the Website Monitor and Downloader Wysigot, Montastic or the Server Monitor Observu.

I have already covered the server monitoring service that I’m using to monitor sites like Ghacks and other websites which offered various packages depending on the amount of servers / websites and response times that you demanded.

Observu is another server monitoring service and it is probably the most basic one that I have ever seen. Users register by entering an email address and password and can log into the website immediately.

The only interaction that is offered is to add a new server to monitor or edit / delete previous entries that are already monitored.

Adding a new website for server monitoring requires minimal user input. Just enter the url of the site and a text that usually appears on it. If that text does not appear during checks an email will be send to the email account of the user.

This limits Observu obviously. It can only detect if a text on a website is not shown anymore. You could write your own script that checks the various services running on your server but that’s probably not what this basic server monitoring service is all about.

It’s great for webmasters who run static pages or pages that rely on MySQL but anything beyond that is more or less out of reach of the server monitoring service.