Inactive Agent¶
Use the following troubleshooting steps if you feel that an agent has become inactive or is not sending events to Observability.
Details¶
DataKitchen provides ready-made agents so you can monitor components in your data estate without making modifications to your existing setup. Each integration agent actively retrieves or listens to events from a specific tool, then forwards the events to Observability.
You can monitor the activity of your deployed agents from the Integrations page.
- The heartbeat provides a timestamp of the last signal received from the agent. A red circle displays with the heartbeat when a signal hasn't been received within the last 4 connectivity check intervals. The interval is set to 5 minutes by default and can be changed on the Project Settings page.
- The last reported event includes the event time and should align with your expectations of work performed in the tool.
Potential causes¶
The causes of an agent experiencing downtime may include:
- An issue with supporting infrastructure.
- The agent's Docker container has stopped.
- The agent's Docker container is unable to monitor the tool.
- The tool being monitored is not working as expected.
Solutions¶
An issue with supporting infrastructure¶
- Check that the service account (SA) API key associated with the agent is valid. An agent with an expired SA key or SA key without adequate permissions cannot send events. Create new keys from the deployment script builder and manage all keys from the API Keys page.
- Check for and address service disruptions in the hosting infrastructure. A Kubernetes cluster or container service that experiences an issue can impact the efficacy of any agents you have deployed.
The agent's Docker container has stopped¶
All Observability agents run as Docker containers and can be deployed anywhere in your data estate that supports Docker. Verify the container is running with:
docker psand/or Docker Desktop , for local deployments.kubectl get pods, for Kubernetes deployments.
The agent's Docker container is unable to monitor the tool¶
If the Observability agent is sending a heartbeat signal but no events, the agent's Docker container may be unable to access your tool due to permissions, networking, or other issues. Check for error logs from the container with:
docker psto note the container ID anddocker logs <CONTAINER ID>, for local deployments.kubectl get podsto note the pod ID andkubectl logs <POD ID>, for Kubernetes deployments.
The monitored tool is not working as expected¶
Observability provides end-to-end visibility across your data operations as work gets done. The absence of events can indicate a potential problem with your tools. When expected events have not occurred for some time, investigate the source and make improvements as needed.
Related topics¶
Observability Integration Agents
Note
Still not working? Contact DataKitchen Support. We can work through this together.