Reliability and user experience are no longer just “nice to have” features. Many companies provide similar services and competition is just a click away. Those with state-of-the-art offerings will be ahead of the pack and have more resources for ongoing improvements. Certainly, speed and uptime are not for free, but the good news is that even with a small budget, reliable applications can be realized.
Early integration of performance assessments, followed by 24/7 monitoring of system health, component, and end-to-end response times on production stages are the pillars for highly available and responsive applications. No matter the size of your company, you should have an application monitoring solution in place that collects health metrics and send you alerts if your critical services are down.
Cloud-based Monitoring
For customers who aren’t interested the technical aspects, a SaaS-based monitoring platform is the way to go. In a cloud-based monitoring setting, the setup and integration of your applications is very easy. You order access, configure the monitoring tasks, and immediately start collecting monitoring data. All operational activities such as backups, security updates or installation of new features on this monitoring suite will be done by your SaaS provider. Focus all your efforts on configuring monitoring tasks, activating alerts, and analyzing results instead of dealing with time consuming deployments. In an on-premises monitoring environment, infrastructure setup, server deployment, firewalls, and other security concerns can easily lead to higher costs and effort as compared to a cloud-based monitoring solution.
Your users are located around the world, and your monitoring approach should reflect this global distribution. A cloud-based monitoring suite comes with a global user simulation network, which is a huge benefit for your monitoring strategy. From a financial perspective, SaaS-based platforms are usage-based. You can start with a small budget and the costs are always in line with your benefits. There’s no need for initial investments.
- The main characteristics of SaaS-based monitoring suites include the following:
- Installed on the providers infrastructure
- Quick deployment
- Zero operational effort
- Low on boarding effort
- Stay up to date with latest product versions
- No maintenance costs
- Pay as you go pricing
On-premises Monitoring Platforms
Before the advent of SaaS-based solutions, monitoring suites were hosted on an organization’s internal network. Organizations agreed on a license contract, payed the initial license fees, and the vendor shipped the setup procedures for their monitoring suites. A dedicated team planned the rollout, ordered the required hardware, installed the monitoring components, and deployed all monitoring agents across their data center. Once the installation was completed, they configured the actual application or infrastructure monitoring devices, set the notifications in place, and educated their users how to use the platform. The internal monitoring team was fully in charge of dealing with all technical monitoring challenges. It’s often a cat and mouse game when it comes to investigation of monitoring issues because a local deployment brings too much uncertainty, delaying issue analysis often for days. In recent years on-premises monitoring suites have become slightly outdated.
The main characteristics of on-premises monitoring suites include the following:
- Installed in your data center
- You are in charge of deployment and maintenance
- Initial license fee
- Annual maintenance fee
- High initial setup effort
Similarities of Cloud-based and On-premises Solutions
Both locally deployed and SaaS-based monitoring suites help to check and improve the health of your applications. Some activities such as configuration, alerting, and reporting are very similar for both deployments.
| Activity | SaaS-based & Locally deployed solutions |
| Configuration | Customers onboard their applications and servers.
The monitoring suite collects health metrics. |
| Alerting | Customers enable the required incident notification.
The monitoring platform notifies your teams if specified thresholds have been exceeded. |
| Reporting | Customers adjust their preferred reporting options.
The monitoring suite generates specified reports. Customers review reports and start a deep dive analysis in order to identify the root cause of identified issues. |
Differences between Cloud-based and On-premises Solutions
Initial setup, maintenance, and license costs are totally different for SaaS-based solutions versus locally deployed monitoring suites. Review the table below before you decide whether to go with the former or the latter.
| Activity | SaaS-based | Locally deployed |
| Deployment | In the vendor’s data center | In your data center |
| Maintenance | The vendor’s job | You are in charge. Setup and roll out of updates can be a huge effort |
| Costs | Pay as you go. You pay exactly what you use. | You pay initial license fee and an annually maintenance. |
| Data storage | In the cloud | In your data center |
Best Practices for Determining the Right Monitoring Deployment
- Clarify your requirements before you choose a monitoring suite. The number of applications in scope, the geographical distribution of your end users, and uptime requirements are crucial elements of your monitoring strategy.
- Monitoring of all layers. It’s not enough to check the health of your back-end services. Your users expect end-to-end availability. Monitor your applications as they are used by your customers.
- Execute end-to-end monitoring from your customers’ locations. Performance and availability can be highly dependent on the point of presence. Network parameters such as bandwidth, round-trip time, and packet loss can have a huge impact on the end-to-end response times of applications.
- The costs to deploy and maintain your local monitoring suites can be high. If you’re not interested in spending more than 30 percent of your monitoring budget on operational efforts, consider a SaaS-based monitoring platform.
- Prioritize your monitoring activities. Start with testing of real use cases and switch on alerting channels. Once this is in place, add service checks and collect system resource utilization metrics.
Tom DeMarco, a well-known software engineer, once said, “You can’t control what you can’t measure.” If you’re currently using antiquated application and website monitoring tools, it’s highly recommended that you consider switching to a forward-thinking monitoring strategy, one that includes monitoring at all layers, continuously checks uptime, and gives you more time for improvement and innovation.
