> For the complete documentation index, see [llms.txt](https://developer.harness.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.harness.io/continuous-delivery/troubleshooting-and-resources/armory/general/debug-armory-io-is-down-crashes-environments-w-debugging-enabled-into-a-crash-loop-restart.md).

# debug.armory.io is Down, Crashes Environments w/ Debugging Enabled into a crash loop restart

### Issue <a href="#issue" id="issue"></a>

Note: This article is for internal use only debug.armory.io becomes unavailable and as a result, Crashloop Restarts will happen for Terraformer and Dinghy. As an example, Dinghy was failing to startup after a restart hence not usable The Full RCA for the incident is found here: <https://armory.slab.com/posts/5-16-2021-debug-armory-io-rca-internal-777v0u67>

### Cause <a href="#cause" id="cause"></a>

armory-kube GKE cluster health for debug.armory.io was not in a good place. As per Christos' investigation

* Node pool workers are not able to join the Kubernetes control plane. Both the control plane and the node pool is running in 1.17.17-gke.4900
* Unable to issue any kubectl commands; Failing with x509 cert error for the Control plane certificates.