Organizations depend on fully functioning IT systems and processes to attract customers, deliver services and manage internal operations. The operation of these systems and processes directly impacts the business and reputation of each organization. Ensuring that all IT systems and processes are fully functional 24×7 is a key component of any organization’s IT performance management strategy. This is where synthetic monitoring comes in.
Synthetic Monitoring Ensures Monitoring of Ongoing IT Service Performance
Synthetic monitoring tracks simulated user transactions so that problems and issues can be quickly discovered and corrected, ideally before users notice and complain. By proactively monitoring IT services end-to-end, synthetic monitoring delivers insight into the performance of the entire service delivery chain — i.e., every IT infrastructure tier supporting the service. IT managers rely on synthetic monitoring to learn about issues before users call in and can, therefore, proactively correct problems before end users are impacted.
Monitoring all the different IT services (whether they are web-based, client-server, or thin-client based) and their access from internal and external locations is a big job! An effective synthetic monitoring tool simulates different transactions to accurately measure service performance (i.e., availability and response time). These user experience metrics reported by synthetic monitoring are indicative of times when the performance of one or more of the IT tiers supporting the service has degraded or when there is an unusual load on the service. Knowing about user experience issues before they become problems is key to efficiently managing the IT infrastructure.
What Organizations Need Synthetic Monitoring?
How much does an online retailer lose if their external web marketplace goes down for even a minute? How much productivity is lost when a payroll or other important internal service fails or is extremely slow? The answers are different for each organization, but the costs are significant. In case the services are being accessed by users external to the organization, the cost of downtime or slow time is even higher. 100% uptime is the objective for all companies, but even a small percentage of downtime or slow time can be incredibly expensive and damaging to the organization’s brand. All types of firms, from eCommerce to manufacturing, need an accurate synthetic monitoring system to ensure efficient operations.
The Real Difference Between Synthetic and Real User Monitoring
While real user monitoring is needed to understand the experience being delivered to real users, unfortunately, when it reports a problem, the issue is already impacting live users. On the other hand, synthetic monitoring runs 24×7, even when there are no real users actively using IT services. Therefore, it can detect problems even during off-hours and allows IT management to work on problems prior to them disrupting the business.
Another point to consider is that synthetic monitoring is performed from well-defined points in the network. Since the target that is measuring the user experience stays constant, any deviations observed in user experience are reflective of changes in the IT service’s performance. On the other hand, with RUM, the endpoints can be different because different users may access the service from different locations and using different desktops or mobile devices. The observed performance may vary depending on the user location and endpoint. Therefore, synthetic monitoring provides a more reliable way of tracking user experience changes for IT services.
Choosing between Synthetic Monitoring and System Monitoring
Many organizations deploy agents on all the servers and desktops in the infrastructure to monitor resource usage levels — CPU utilization, disk activity, network bandwidth used, etc. Just because a server has low CPU utilization does not mean that the response time of the service is going to below. Therefore, monitoring IT systems using agents is not sufficient for ensuring a good user experience.
Synthetic monitoring is complementary to system monitoring. When synthetic monitoring identifies a problem with user experience, system monitoring may provide diagnostics that highlight why the problem happened (i.e., a backup job on a database server can slow down a web application).
In today’s economy where user experience is the primary way of measuring IT performance and where IT budgets are tight, synthetic monitoring is the most effective, easy and cost-effective way to monitor the performance of your IT systems, processes and services.
Synthetic Monitoring is a Must for Cloud or SaaS services
These days, most organizations are relying on one or more cloud-based or SaaS services. With a cloud-based or SaaS service, organizations are able to access and consume the service, but there is no way to install agents at the service provider end. Even though the cloud or SaaS service provider may provide monitoring consoles and APIs, these do not provide an unbiased way to monitor the performance of the service. Synthetic monitoring is the only way to monitor the performance of cloud and SaaS services in an unbiased manner and the results of synthetic monitoring can be used to measure performance against promised service levels.
Many organizations also have hybrid IT services where the main business logic is executed on systems internal to the network, but they use external, third-party services for specialized functions.
For instance, an eCommerce web site uses a third-party payment processing gateway. As in the case of cloud/SaaS services, synthetic monitoring is the best way of measuring the performance of the hybrid IT service as a whole, or the third-party payment processing gateway individually. With synthetic monitoring in place, IT managers can be made aware of potential issues and problems concerning third-party services. This helps them to ensure that their third-party service providers are delivering the performance expected of them.
Synthetic Monitoring is Key to IT and ROI Planning
Another important use of synthetic monitoring is for baselining the performance of IT services. This is particularly handy when there are changes in the offing. For instance, if you plan to upgrade your core application, baseline service performance before the upgrade and after the upgrade using synthetic monitoring. The comparison can indicate how well the upgrade went, or if there are issues following the upgrade, you will know of them and can take action immediately. Many IT managers adopt synthetic monitoring and baselining as a best practice during software upgrades, migrations from on-premise to the cloud, adoption of SaaS services, etc.
Diagnosing the Cause of Problems Using Synthetic Monitoring
Since it adopts a black-box view of the IT service it monitors, synthetic monitoring is more useful for detecting problems than for the diagnosis of problems. At the same time, by strategically performing synthetic monitoring from different locations, IT managers can narrow down the cause of problems. For instance, by comparing the response time reported by a synthetic monitor placed in the data center itself to the response time reported from a distant branch, an IT manager can determine if the problem is specific to a branch or is affecting all the branches (i.e., because the response time in the data center itself is high).
A well-structured synthetic monitoring system provides not only the alert but delivers important information about location and particulars of the issue so that managers need not waste time trying to find the problem.
Synthetic Monitoring is NOT the same as Stress Testing
In some sense, synthetic monitoring and stress testing are similar. They both simulate user interactions with IT services. The difference lies in the number of simulations performed. While synthetic monitoring assesses performance by simulating one user, stress testing is focused on simulating hundreds or thousands of users, so it can determine if the target infrastructure can withstand this workload or not. Since it imposes a significant load on the target infrastructure, stress testing is often performed during off-peak hours or weekends, or in staging environments. On the other hand, synthetic monitoring does not stress the target infrastructure — the workload only sees the addition of one user. Hence, its impact on production services is negligible. Synthetic monitoring is therefore well suited for use in production environments.
How eG Enterprise Can Help
eG Enterprise provides state-of-the-art synthetic monitoring solutions for IT teams to proactively test, detect and diagnose problems. Administrators can choose from a variety of synthetic monitoring functionalities — logon simulation of virtual applications and desktops, full session simulation for VDI sessions and thick client apps, web app simulation and more. IT teams and application owners can use synthetic monitoring to baseline the performance and user experience and compare across locations to identify deviations.