Application uptime and performance monitoring needs are different in the DevOps age. So is the approach that MSPs in managed IT services should take to monitoring.
Brought to you by MSPmentor
If you're performing software monitoring in the same way you did ten years ago, your monitoring routine is out of date.
Here's how monitoring has changed as a result of the DevOps revolution and what it means for the managed services industry.
Traditionally, the main focus of monitoring was preventing downtime.
Monitoring tools were essentially used to determine when a server or application went down.
More sophisticated tools would perform health checks to tell you when something looked wrong with a server or app so that you could fix it before it crashed.
Monitoring in the DevOps Age
Today, effective software monitoring requires much more than a focus on maintaining uptime.
DevOps and the technologies and practices that have come with it -- like containers and microservices -- have ushered in new goals and requirements for monitoring. Consider the following points:
· Optimizing cost has become as important to many organizations as optimizing performance. The cloud, serverless computing and highly scalable infrastructure make it much easier today than it was in the past for organizations to consume only what they need and avoid paying for more. For that reason, part of the purpose of monitoring tools is to help identify underutilized resources in order to save money.
· In containerized microservices environments, containers and servers can go down by design whenever demand decreases. This makes it futile to focus on monitoring containers or servers for downtime. Your focus needs to shift to monitoring microservices themselves, rather than the infrastructure that hosts them.
· Containers and microservices also mean that there is much more to monitor. A decade ago, even a complex app deployment usually involved only a few servers. Today, your containerized environment could include thousands of containers spread across dozens of physical or virtual servers -- not to mention layers of software-defined storage and networking. This additional complexity dramatically increases the number of objects you have to monitor.
· Monitoring is no longer just the job of the IT department. Thanks to DevOps's emphasis on flexible roles, plus the growing importance of regulatory compliance, it's crucial to be able to share monitoring data in an access-appropriate way across the organization.
Monitoring and MSPs
What does all of the above mean for MSPs?
Essentially, that their approach to monitoring needs to be more flexible, more sophisticated and more comprehensive than it was in the past.
Whether you monitor your clients' infrastructure as part of their service plan, or just monitor the in-house infrastructure that you use, you need to think in terms of more than just uptime and downtime.
Today, it's important to think also about cost optimization, the accessibility of information, compliance and more.
You also need to make sure your monitoring tools and workflows are capable of handling next-generation infrastructure like microservices.
While most legacy monitoring tools can be extended or adapted to work with Docker containers, few can support containerized environments well out-of-the-box.
They need to be reconfigured to focus on monitoring services rather than infrastructure, and to handle the complexity of large, agile containerized environments.