Linode favicon

Linode outage on 2026-10-20

linode.com
October 2026 20 Minor incident

If you tried to provision a new LKE Enterprise cluster on 20 October, you likely hit a wall for just over nine hours. The fault sat with Linode's cluster provisioning pipeline, not with anything in your own configuration.

Started

12:29 UTC

Duration

Lasted 9h 19m

Source

IsDown

Next time

Linode is degraded. Is your product working?

An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.

50 monitors for your own websites. No card.

Affected components: US-West (Fremont) Linode Kubernetes Engine, CA-Central (Toronto) Linode Kubernetes Engine, EU-Central (Frankfurt) Linode Kubernetes Engine, AP-South (Singapore) Linode Kubernetes Engine, AP-West (Mumbai) Linode Kubernetes Engine, AP-Southeast (Sydney) Linode Kubernetes Engine, AP-Northeast-2 (Tokyo 2) Linode Kubernetes Engine, US-East (Newark) Linode Kubernetes Engine, US-Central (Dallas) Linode Kubernetes Engine, EU-West (London) Linode Kubernetes Engine

What happened?

Linode's status page opened a single incident at 12:29 UTC affecting LKE Enterprise cluster creation across all regions. Ten components were listed as affected, covering every LKE Enterprise region Linode runs: Newark, Dallas, Fremont, Toronto, London, Frankfurt, Singapore, Tokyo 2, Mumbai and Sydney. The description was specific: new nodes could reach a "Ready" state, but system critical components could fail to start, which meant workloads would not run even though the node itself looked healthy. Linode posted a short update four minutes later saying they were continuing to work on a fix, then repeated the original investigating message at 12:34. At 14:06 Linode said the issue had been identified and a fix was being implemented, though no further detail on the root cause was given at that point or later. The incident stayed open until 21:42, when Linode marked it resolved, stating they had not observed any additional issues with the LKE service. That is a gap of roughly seven and a half hours between "identified" and "resolved" with no update in between, so for that stretch you had no way of knowing from Linode's page alone how close a fix actually was. No user reports were filed during the window on this page. Linode's own written record is the only account of what happened, and it names the symptom but not the cause.

Learning

This incident sat specifically in cluster provisioning, the step where a new LKE Enterprise cluster gets created and its system critical components start up, rather than in an already running cluster's data plane. That distinction matters because a standard uptime check against an existing, already-provisioned cluster would have stayed green the whole time, while anyone trying to create a new cluster during those nine hours and nineteen minutes would have hit a broken deployment regardless of what any dashboard said. Linode's 90 day history shows 20 incidents with a median duration of 5 hours 45 minutes, and this one ran longer than that median. Over the last 30 days, 10 incidents were recorded with a median duration of 9h 19m, so a wait of this length is consistent with what Linode has been producing lately, not an outlier. If your own monitoring only watches endpoints you control, it cannot see a failure in Linode's provisioning pipeline itself, because there is nothing of yours to check until provisioning finishes. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.

4.7
stars out of 5
284+ reviews on

Start monitoring in 30 seconds.

There's nothing to install. No credit card required. 50 monitors for free.