Brief internal cluster DNS failure in US-EAST-4


At 4:49 PM UTC, the physical server that housed the US-EAST-4's internal DNS server became briefly unresponsive, which temporarily prevented project instances from communicating to one another. Our systems automatically migrated the DNS server to a healthy physical server a few seconds later.

To prevent this situation from happening in the future, we're adding additional redundancy to our DNS server setup, so that should one DNS server fail, there are others to facilitate the communication between instances.

We're also applying an upgrade from our hosting provider to the physical servers in the cluster to prevent the same kind of failures from reoccuring.

Avatar for
Began at:

Affected components
  • Clusters
    • US-EAST-4