Starting at 4:02 PM ET on October 7, 2018, a small portion of the pages in our customer portal were unavailable. Service was incrementally restored, and fully available by 9:14 PM ET.
Consumer facing services were unaffected. Underlying data was unaffected.
During routine patching of systems, a portion of our compute nodes experienced failures in their network interfaces. These issues were not seen when applying the same patches to our pre-production environments. We were able to move the workloads on these nodes to other nodes to mitigate the issue and restore the full environment shortly after.
We are adjusting our patching process to make some of our pre-production environments more closely match production configurations and load.