From the course: MLOps Essentials: Model Deployment and Monitoring

Unlock the full course today

Join today to access over 24,500 courses taught by industry experts.

Instrumentation for observability

Instrumentation for observability

- [Instructor] Instrumentation is the first vital step to ensuring a good foundation for observability. Instrumentation refers to adding capabilities to systems and applications to track and capture information that can be used to observe the behavior and performance. Instrumentation deals with adding hooks into software that enables this tracking. Let's begin with instrumentation at the system level. What kind of data is captured here? We want to track CPU usage for the container and by individual executables. Similarly, we want to track memory usage. Disk I/O stats are essential to understand storage load. Network traffic information helps understand the performance of the network. Request/response statistics are collected in API gateways to understand the incoming requests and their responses. Error statistics at the system level will monitor any exceptions with resource access and handling. To capture this…

Contents