AWS recently added the eu-west-2d Availability Zone to the London region, but the announcement only appeared publicly after the change. Some customers reportedly experienced problems afterward. Did anyone see deployment failures, networking issues, or other service disruptions related to the new Availability Zone?
4 Answers
Our Terraform deployments became difficult for some teams after the change. Configurations had to be updated, and a few deployments failed partway through. The new zone also appeared to lack support for some commonly used services, including PrivateLink and Transit Gateway, so simply treating it like a fully equivalent zone caused problems.
There was also a reported EC2 service disruption in eu-west-2 around the same time, and it was later resolved. That could explain some of the network and instance failures people noticed, although it may have been separate from the general rollout of the new zone.
The broader concern is capacity and redundancy. London is still only one AWS region, and the individual Availability Zones are relatively small. Adding another zone increases capacity, but it does not provide the same protection as having a second geographically separate UK region. Customers needing UK-only data residency may still have limited disaster-recovery options.
A likely failure mode is code or infrastructure configuration that assumed there were only three Availability Zones. Anything iterating over a fixed A/B/C list could suddenly encounter a fourth zone and recreate resources or fail during deployment. It may not be an AWS outage so much as an assumption in customer automation being exposed.
That distinction makes sense—having the zone available for EC2 does not necessarily mean every regional service supports it immediately.