Deployment logs#

When you “deploy” your project, Data Science & AI Workbench creates various Kubernetes resources, including a Kubernetes Deployment, which manages the lifecycle of pods—including their creation, scaling, and updating as necessary. Each project deployment is created as a Kubernetes pod.

Each pod has an app and proxy container, and logs for each container are accessible to assist you in troubleshooting issues that can occur within your project’s deployment. These logs can provide insights into the operational health of your deployment, possible configuration errors, and detecting potential security issues. You can also access Kubernetes event logs to aid you in troubleshooting your cluster.

Viewing deployment logs#

To view a deployment’s logs:

  1. Log in to Workbench.

  2. Open a project.

  3. Select Deployments from the left-hand navigation.

  4. If necessary, deploy your project.

  5. Select your project deployment.

  6. Select Logs from the left-hand navigation.

  7. Open a container dropdown to view the logs for that container.

Tip

Use the search function to help you efficiently locate information in the container logs.

App container logs#

The app container is where your project application code runs. Logs from this container can provide insights into the following:

  • Standard output (stdout) and standard error (stderr) from your application

  • Error messages and stack traces when exceptions occur

  • Information related to the applications resource usage and processing times

Proxy container logs#

The proxy container handles network traffic to the application from both the internet and from other internal Kubernetes services. Logs from this container can provide insights into the following:

  • Records of incoming and outgoing requests, including timestamps, HTTP status codes, URLs, and client IP addresses

  • Records of unauthorized access attempts to help identify and address irregularities in network traffic

  • Information about network-related errors, such as failed connections, timeouts, or DNS issues

Events logs#

The events logs contain actions taken by Kubernetes as it schedules pods for creation on the node (as they consume cluster resources), pulls the image, and then starts each of the pod’s app and proxy containers. These logs provide insights into the following:

  • Information about the automatic scaling actions taken by Kubernetes in response to workload demands

  • Records of new version deployments, configuration updates, and their success or failure

  • Results of health-check events to ensure your application is healthy