In the modern world, businesses are heavily reliant on IT systems for their customer interactions, sales and internal business processes. Any outage can be extremely costly to a company’s reputation and its bottom line.
This article outlines how different type of IT monitoring solutions can be applied to reduce the impact of outages as well as exploring the type of systems that may right for your business Specifically, we’ll cover the broad differences differences between application monitoring, synthetic monitoring and real user monitoring.
The cost of critical IT outages.
In a 2017 Statista survey of more than 750 companies (December 2017), it was found that the average cost of a critical IT outages was between US$300,000 and US$400,000. Even at the lower end, the average cost came in at over US$100,000 per hour. A Gartner survey (July 2014) concurs, also estimating the cost of an outage at being over US$300,000 per hour.
Outside of the direct cost, outages for any customer facing system can have a huge impact on the reputation of a business.
How IT monitoring can help.
We all know that IT monitoring is an important core service to help avoid outages and reduce the time of any outages that do occur. But not everyone is familiar with the subtle differences between different types of monitoring.
At its most basic, IT monitoring involves establishing an automated system to constantly check on the entire IT operation to ensure that it is running as expected. When any part of the system starts to run into trouble (for example, a website is unavailable), IT operations staff can be immediately notified and begin the work of getting the systems up and running again.
The effect of IT monitoring is two-fold – it ensures that the business knows about issues, preferably before customers notice. And it can also help pinpoint the exact area (server, network component, etc) that needs to be fixed.
With the cost of critical outages being so high, IT monitoring is a critical tool to help reduce the time to resolve the problem.
A common starting point for IT monitoring is a bottom-up approach where the health of each individual server and network component is monitored. This information is vital to help pinpoint the possible cause of issues.
But knowing how each component is running is not enough. Given that an application can be distributed over multiple servers in multiple locations, it’s vital to use application monitoring to check how each application itself is running.
There are two main approaches to application monitoring – synthetic monitoring and t real user monitoring. They each perform important roles but in uniquely different ways.
Real user monitoring.
As the name suggests, real user monitoring (also known as RUM) monitors the user experience in real time. This is typically implemented by adding a monitoring agent to an application, such as by adding a java script tracking mechanism into a web-based application.
Real User Monitoring can be used to track user movements through an application, recording how long a page takes to load, how long they spend there, and how they move around the application.
The main advantage of Real User Monitoring is that it gives you the actual information about what is happening. This can give a clear indication of what users are doing as well as help to check that service levels (such as page load times) are being met.
Its main disadvantage is that you only find out about a problem when users are already experiencing it. What’s more, it can also be difficult to deploy on legacy applications or applications run by an external party (such as Gmail).
Synthetic monitoring.
Synthetic monitoring is used to proactively test common user journeys through applications to identify issues before they affect real users. Automated tests are set up to try out the most critical user journeys.
For example, a synthetic test could login, search for a product, make a purchase and logout.
The tests can be run as often as needed.
Synthetic monitoring can also be used to test changes to the application before the changes are released to the public – making sure that the updates haven’t broken the existing functionality.
The main advantage of Synthetic Monitoring is that any issues can be discovered before end users notice it. In the above example, if there was an issue with the website checkout process, the test should uncover it quickly – hopefully before sales are lost.
Synthetic monitoring is excellent at providing clear metrics around application performance. By running the test in a controlled environment, it can eliminate variations due to everyday issues like varying network performance as users are on the move. This can help you focus on the areas that need to be investigated when issues are highlighted.
Its main disadvantage is that the tests need to be updated every time the applications changes. For instance, a user interface update could potentially upset existing test scripts.
Synthetic Monitoring products need to be carefully selected to ensure they can deliver the function your IT infrastructure requires. Not all products are created equal. If the wrong product is selected, it can take a lot of time and technical expertise to initially setup as well as continuing to maintain test cases.
What type of IT monitoring is best for your business?
As you would expect, an effective IT monitoring strategy uses all of the above:
- real-time user monitoring to identify application issues that are affecting users
- synthetic monitoring to monitor application performance and try to avoid users seeing any issues in the first place, and
- monitoring of each individual server and network component to help diagnose any problems and deal with common issues before they impact on the user experience.
The 2Steps synthetic monitoring solution.
Until now, synthetic monitoring has been difficult to get right. Traditional monitoring tools have been limited in what they can test, and have taken significant technical skill and time to create and deploy.
2Steps changes all this. It’s extremely easy to setup, works with any platform and is cost effective. It’s a true ‘no code’ solution that reduces your dependence of engineering resources.
With its simple point and click user interface, you can create new test cases to test any application in minutes.
2 Steps will work with anything – applications that are web-based, windows applications, Virtual Desktop environments like Citrix, AVD and Omnissa Horizon. It even works with java applications, mobile apps and systems that require 2-factor authentication.
With 2Steps, you can utilise machine learning to link different types of monitoring data together. This creates powerful insights for your operational support team, helping you to resolve outages faster than ever before.
2Steps brings together a suite of highly useful functions, making it the go-to synthetic monitoring solution.
- Lerner, A. (2014, July 16). “The Cost of Downtime – The Gartner Blog Network”. Retrieved from https://blogs.gartner.com/andrew-lerner/2014/07/16/the-cost-of-downtime/
- Statista (2017, December). “Average cost per hour of enterprise server downtime worldwide in 2017 and 2018”. Retrieved from https://www.statista.com/statistics/753938/worldwide-enterprise-server-hourly-downtime-cost/
