[xxdesmus]

As the blog post outlines, the outage was related to an upstream network provider leaking routes. Note exactly something we can prevent for them.