Blocks API deploys took an abnormally long time because of a pre-existing dependency between a few internal services which were all simultaneously updated between 1:20 and 1:30 PM pacific time today.

We've rolled out a change which decouples these services so that these 502s won't re-occur in the future.

This automated health-check failure resulted due 502s from our blocks.dopt.com gateway due to an abnormally slow roll-out of new nodes in k8s.

We're investigating why this roll-out was so slow, but all systems are operational and functioning as expected.

A previously reported automated failure on health-check has been resolved.