Google VM, microservices tools advance cloud migration strategy

Google hopes to advance its third-place standing in the public cloud market with additional cloud migration tools and enterprise-friendly features to entice more workloads onto the platform.

Migrate for Compute Engine can now move virtual machines from Microsoft Azure to Google Cloud, although the feature is in beta. Previously, Migrate for Compute Engine supported only migrations from AWS.

Google has also expanded support for service mesh, which provides a communication layer for application components to talk to one another. It’s favored for microservices-based application architectures and relieves the burden on developers to inject and maintain networking code in their apps.

Traffic Director, Google’s take on a service mesh control plane that was unveiled at Cloud Next in April, is now generally available, according to a blog post. Google terms Traffic Director as a global traffic manager for VMs and container workloads. It is now available for use with Anthos, the Kubernetes-based multi-cloud container management platform that became generally available in April.

Traffic Director provides a Google-managed version of Pilot, the traffic-management component of Istio, a popular open source service mesh launched by Google, IBM and Lyft in 2017. Pilot manages traffic between microservices on the network through the Envoy sidecar proxy, which executes the distributed networking functionality for apps.

If I’m just giving you a container platform that can run anywhere, there are a lot of companies that can do that with Kubernetes.
Gary ChenAnalyst, IDC

Google’s Layer 7 Internal Load Balancer is now in beta and based on Traffic Director and Envoy. It will give users comprehensive traffic control capabilities with the feel of a traditional load balancer, ideal for projects that involve legacy app migration to service mesh, according to Google.

Anthos is Google’s long-term multi-cloud play

Google has positioned Anthos as an ideal platform for companies that want to achieve parity for container workloads across multiple clouds and their own data centers. Anthos, which is built on Kubernetes, gives customers the ability to refactor applications once and then run them anywhere, according to Google.

While Google is happy to make money from customers that move VM workloads to Compute Engine, it also sees an opportunity for Anthos as a platform for application modernization. Companies that convert VMs to containers via Anthos Migrate get benefits, such as no more need to manually patch their OS, Google said in the blog.

Anthos is aimed at large enterprises and is priced accordingly. One clear rival is Red Hat’s OpenShift, which has similar intent around cross-platform workload portability, albeit with some technical differences. Now that the IBM acquisition of Red Hat is closed, Big Blue likely will unleash its massive global sales force to support OpenShift.

Gary ChenGary Chen

Although Google said it plans to offer Anthos on other public clouds, specific timelines are unclear. “I’m taking more of a wait-and-see attitude to see how that really turns out,” said Gary Chen, an analyst with IDC. Anthos is strongly tied to the Google Cloud, and its true value may be its association and integrations with other Google Cloud services, such as for big data analytics and machine learning.

“If I’m just giving you a container platform that can run anywhere, there are a lot of companies that can do that with Kubernetes,” Chen said.

Even as Google seeks to convert VM workloads to containers for Anthos, it has acknowledged a need to work with VMWare on other fronts. This week, the companies announced the upcoming availability of a VMWare on Google Cloud service.

The move followed the recent release of a similar service on Azure, while VMWare Cloud on AWS has been available since 2017 and gained significant market traction.

Ultimately, enterprise customers have more choices to port VMware workloads onto the public cloud, but they must now weigh considerations such as relative costs and the strength of associated partner ecosystems.

Go to Original Article
Author: