From the course: DevOps for Data Scientists

Unlock the full course today

Join today to access over 22,600 courses taught by industry experts or purchase this course individually.

Monitoring models in production

Monitoring models in production

From the course: DevOps for Data Scientists

Start my 1-month free trial

Monitoring models in production

“

- [Instructor] DevOps is about developing systems and keeping them running. Performance monitoring is an essential practice to meet that objective. Performance monitoring spans a wide range of tasks, including understanding the level of resource utilization, knowing what services are available to users, measuring throughput and operating efficiently, and tracking model quality, especially when production data may change over time, leading to the need to retrain models on newer, more representative data. Monitoring resources is about making sure key resources of CPU, memory, and persistent storage are sufficient for your application needs. There are many GUI tools for monitoring and visualizing the status of resources. And command line tools are also informative and easily combined in strips to collect and display information targeted to your particular interests. On Linux, some popular command line tools include top for CPU monitoring, free for memory monitoring, df and du for disk…

Contents