Troubleshooting

Troubleshoot common env zero self-hosted Kubernetes agent issues, including SIGKILL and SIGABRT errors caused by pod memory limits or the Node.js heap limit.

Deploy failed with signal SIGKILL or exit code 137

If your deployment failed with deploy failed with signal SIGKILL or exit code 137, the deployment pod most likely ran out of memory. When a container exceeds its memory limit, Kubernetes kills it with SIGKILL and marks the pod OOMKilled. Run kubectl describe pod <pod-name> and check the last state’s reason to confirm.

In this case, define requests and limits for the pod size as described in Custom/optional configuration.

Deploy failed with signal SIGABRT

If your deployment failed with deploy failed with signal SIGABRT, the deployment process most likely ran out of Node.js heap memory. The pod is not OOMKilled in this case, so kubectl describe pod doesn’t show it. Run kubectl logs <pod-name> and look for Reached heap limit or JavaScript heap out of memory to confirm.

Raising the pod’s memory limits helps only if the pod limit, not the Node.js heap limit, is the one being reached.

Next steps

• Check the agent’s status
• Upgrade the agent release after changing its resource values