IL-0156 Linode · 2025-09-24kubernetes
The certificates on LKE cluster controllers were about to expire and were rotated on 23 September 2025. As part of the follow-up work after that rollout, the old certificates on the associated controllers were revoked at 02:29 UTC on 24 September, and the impact began right after. The investigation found a certificate key mismatch between the API endpoints and the cluster controller, which led to the 5xx errors on cluster management requests and the impact on autoscaling.
Linode deployed a fix and then restarted the API endpoints; the impact ended at 06:54 UTC, about four and a half hours after the revocation. The status page moved to monitoring at 07:23 and marked the incident resolved at 08:05. In the postmortem posted at 17:18 UTC, Akamai's subject matter experts were still investigating the root cause.
From about 02:29 UTC on 24 September 2025, customers of Linode Kubernetes Engine (LKE) in US-Southeast (Atlanta) could not manage their clusters, and autoscaling was impacted. Internal logs showed 5xx errors whenever customers tried to manage LKE clusters associated with the affected cluster controller.
Only cluster management was affected: changing node pools, scaling, and creating or deleting clusters. Running customer workloads kept working.
The status page opened the incident at 04:31 UTC as an emerging issue under investigation, posted another update at 05:42, and reported the root cause identified at 06:28.
Symptom ? High error rateAutoscaler misbehaviour
Root cause ? Certificate or credential expiry
–
–
–