Mathias Rosenthal - Fotolia

Get started Bring yourself up to speed with our introductory content.

AWS App Mesh faces tough race in microservices market

As AWS and its rivals continue to evolve their service mesh technologies, they'll have to do so with enterprise demands, including those for multi-cloud capabilities, top of mind.

To help enterprises tighten their control over containers and microservices architectures, AWS rolled out App Mesh at re:Invent 2018. This year, it will have to keep pace with its rivals in the service mesh market and try to align its offering with enterprises' multi-cloud needs.

AWS App Mesh marked the cloud provider's official entry into the service mesh technology fray. Because microservices architectures can involve hundreds of services, they make it more difficult for IT teams to establish visibility into, or troubleshoot, an environment, compared to a traditional, monolithic architecture, said Fintan Ryan, a Gartner analyst.

AWS continues to implement App Mesh, which uses the open source proxy Envoy, across several of its services, including Amazon Elastic Container Service, Amazon Elastic Container Service for Kubernetes and AWS Fargate.

Service mesh targets both dev and ops needs

A service mesh is comprised of two layers: the control plane and the data plane. The control plane is the management layer, and the data plane is where different services communicate. Envoy operates as a sidecar proxy in the data plane and passes traffic to appropriate destinations.

The service mesh approach has emerged in response to "cross-cutting concerns," Ryan said. This issue develops when different parts of a system become intertwined but still have to meet specific requirements. The service mesh sidecar model handles those requirements more efficiently, as Envoy instantiates a unique identifier for a given process or activity, he said. This enables capabilities such as distributed tracing, observability, traffic routing and traffic shaping.

"[Envoy] is a way to deal with layer seven -- the application layer [in the Open Systems Interconnection model] -- so that you don't have to worry about what is going on below that," Ryan said.

This will appeal to both developers and operations teams who use AWS. For instance, App Mesh will help developers minimize these cross-cutting concerns and maintain logs of microservices communications. Meanwhile, the service provides operations teams with a consistent way to enforce policies across microservices architectures.

"Developers will care more about what is happening on the data plane and operators will be interested in setting policies," Ryan said.

The sidecar model is attractive for both infrastructure and development teams, agreed Jeffrey Hammond, a Forrester analyst.

"We've seen good results with microservice[s] architectures that use sidecars, like the Envoy proxy, to standardize communication between services," he said.

The primary use case for AWS App Mesh is to standardize and scale microservices architectures, Hammond said -- something that companies like Lyft or Stripe have done to expand their infrastructures. However, many enterprises don't necessarily want to manage their own Kubernetes and service mesh environments.

"They don't have the skills, expertise or time to set up, configure and run them," Hammond said. That's where App Mesh, as well as Amazon Elastic Container Service for Kubernetes, come in -- as a managed service for organizations that want to get up and running more quickly.

How the service mesh race will shake out

Just as 2018 was the year of the managed Kubernetes service, 2019 may be the year of service mesh, Hammond said. And AWS competitors have also taken notice.

Microsoft has Azure Service Fabric in general availability and Google has integrated Istio, an open source service mesh platform, with Google Kubernetes Engine as the basis for its offering, which is in beta. It's still too early, Hammond said, for a detailed breakdown or comparison of these services, especially because they'll likely evolve throughout 2019.

However, the race between these providers is about more than simply having a great service mesh option, said Bob Laliberte, an analyst at Enterprise Strategy Group. For example, his firm's research on 2019 tech spending shows that about 70% of companies polled already had multi-cloud environments. That presents a challenge for each of these providers, as they'll have to help customers operate service mesh technology not only on their own cloud platform, but across other platforms, as well -- a capability that is still rudimentary at best.

Therefore, if a cloud vendor can provide a single service with multi-cloud capabilities, it may gain more customer loyalty. This is likely where AWS and its competitors have set their sights -- and the fact that Envoy is a component in each vendor's offering appears to be a starting point.

Dig Deeper on AWS architecture and design

Join the conversation

1 comment

Send me notifications when other members comment.

Please create a username to comment.

Which service mesh capabilities do you need readily available in your IT role?
Cancel

-ADS BY GOOGLE

SearchCloudApplications

TheServerSide.com

SearchSoftwareQuality

SearchCloudComputing

Close