When operating infrastructure and applications, monitoring is indispensable for gaining an insight into the status of the respective components. This applies not only to local environments, but also to the public cloud. The following article provides an introduction to monitoring the Google Cloud Platform. Among other things, we look at monitoring virtual machines, setting up alerts, using dashboards and setting service level objectives.
With its Google Cloud Platform (GCP), Google is one of the three major public cloud providers alongside AWS and Microsoft Azure. In recent years, Google has seen considerable growth, particularly in the areas of machine learning and big data. However, GCP also has nothing to hide from its competitors in the classic IaaS sector. Monitoring plays a central role in the proper operation of a cloud infrastructure. At Google, the Cloud Operations Suite, which addresses the topics of monitoring, logging, tracing, debugging, profiling and auditing, was called Stackdriver until two years ago. Google no longer uses this name, but simply refers to it as the Operations Suite.
In order to ensure functioning monitoring during operation, it is necessary to receive data from the source systems in the form of signals. In the case of monitoring, these are metrics. These can originate from IaaS components such as virtual machines, from higher-value services such as managed databases, from platforms such as Kubernetes, from microservices, but also from applications themselves. It is also necessary to keep an eye on incidents, whether in the form of alerts, error reports or service level objectives. With the Operations Suite, all this data can be brought together, viewed more closely, visualized and used for troubleshooting.
With GCP, monitoring data can be accessed both centrally and decentrally. In the Google Cloud, all resources are assigned to projects. In the Operations Suite, each project can collect and analyze the data on its own. However, if you want to maintain an overview of all systems across the entire organization, you should add several projects to a monitoring workspace.
