DocsRun Logs
Experiment TrackingRuns DashboardRun Logs

Overview

Polyaxon comes with a powerful and uniform logging experience for all runs.

Polyaxon collects and archives logs from containers in runs, including distributed jobs, experiments, notebooks, and TensorBoard services. A consistent logging view helps diagnose failed jobs and services that are not accessible.

Logs in CLI

You can stream logs from the CLI or the UI to investigate container issues without querying Kubernetes directly.

run-logs-cli

Logs in UI

You can stream logs in the dashboard as well with support for progress bars, ANSI standards, and long log lines.

run-logs-dark

Use the search field to find messages, with optional regular expression and case-sensitive matching. Containers, Pods, and Nodes filter the source of the messages. The example below uses generated documentation logs from acme/demo.

run-logs-filter

Use Columns to choose the source fields to display and Time Format to change timestamp formatting. The log panel also provides controls to jump to the beginning or end and download the logs.

It’s also possible to disable logs storage for a specific experiment or based on a specific tag, e.g. users might want to avoid storing logs for debug jobs.