The Enterprise Path to Service Mesh Architectures, 2nd Edition

Chapter 3. Adoption and Evolutionary Architectures

What practical steps can be taken to adopt a service mesh in my enterprise?

As organizations adopt service mesh architectures, they often do so in a piecemeal fashion, starting at the intersection of the most valuable (to them) feature and the deployment model with lowest risk as calculated based on their operating environment and their current operations skill set.

Piecemeal Adoption

Desperate to gain an understanding of what’s going on across their distributed infrastructure, many organizations seek to benefit from auto-instrumented observability first, taking baby steps in their path to a full service mesh after initial success and operational comfort have been achieved. Stepping into using a service mesh for its ability to provide enhanced observability is a high-value, relatively safe first step. First steps for others might be on a parallel path. A financial organization, for example, might seek improved security with strong identity (assignment of a certificate to each individual service) and strong encryption through mutual TLS between each service, while others might begin with an ingress proxy as their entryway to a larger service mesh deployment.

Consider an organization that has hundreds of existing services running on virtual machines (VMs) external to the service mesh that have little to no service-to-service traffic, rather nearly all of the traffic flows from the client to the service and back to the client. This ...