Linode's US-Southeast (Atlanta) region had a service issue that lasted 48h 11m, starting at 17:49 UTC. Compute instances in that region were unable to boot for the full duration, and the fault sat with Linode's infrastructure, not with anything in your own setup.
Started
17:49 UTC
Duration
Lasted 2d 0h
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-Southeast (Atlanta)
What happened?
Linode's status page opened an incident at 17:49 UTC for US-Southeast (Atlanta), describing an emerging service issue linked to an increasing number of host failures. The practical effect was that compute instances in that region could not boot. Linode posted two further updates, at 19:03 and 20:17 UTC, both saying only that the investigation was continuing, with no cause given and no timeline for a fix. After 20:17 UTC there is nothing further on record from Linode, yet the incident itself ran for 48h 11m before being marked resolved. That means well over a day of the outage passed without any public update describing what was wrong or what had changed. This is a single-component incident: only US-Southeast (Atlanta) is listed as affected, so other Linode regions are not implicated by this record. No user reports were filed during the window, which on a page like this just means nobody logged a report through this channel, not that nobody was affected. For context, Linode's own history shows 27 incidents in the past 90 days with a median duration of 6 hours 31 minutes, so a 48-hour outage is far outside its usual pattern. Over all time Linode has logged 599 incidents, averaging 7.7 a month, so incidents in general are not rare for this provider, but one running two days with three terse updates is unusual even by that standard.
Learning
If your workload runs on Linode's US-Southeast (Atlanta) region, this incident sat at the host and boot layer, below anything your own application monitoring can see. A health check against your own endpoint tells you whether your product is reachable, but it says nothing about whether a replacement instance can actually boot in a degraded region, which is exactly what failed here. UptimeRobot's own 30-day figures for Linode show 16 incidents, with only 11 resolved and a median duration of 19h 9m once resolved, so long, slow-moving incidents like this one are not a one-off for this provider. The shape to watch for is a vendor status page that stalls on "still investigating" for hours at a time while the underlying problem, instance boot failures, keeps affecting anything that needs to scale or restart in that region. Your own monitoring and Linode's status page are both doing their job correctly here: one watches your product, the other watches Linode's infrastructure, and they can disagree without either being wrong. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.