Janakiram MSV Contributor
Enterprise & Cloud
I cover Cloud Computing, Machine Learning, and Internet of Things:
For More Information: Forbes
At the recently concluded Cloud Next conference, Google announced the general availability of Anthos, an enterprise hybrid, and multi-cloud platform.
Though there were over 100 announcements made at the event, Anthos stands out for one apparent reason. It marks Google’s official entry into the enterprise data center. It is one of the first official multi-cloud platforms from a mainstream public cloud provider.
From Sundar Pichai to Thomas Kurian to Urs Hölzle, Anthos was talked about as the next-generation technology. It is evident that the top leadership team at Google is proud of Anthos.
Anthos is different from other public cloud services. It’s not just a product but an umbrella brand for multiple services aligned with the themes of application modernization, cloud migration, hybrid cloud, and multi-cloud management.
Despite the extensive coverage at Google Cloud Next and, of course, the general availability, the Anthos announcement was confusing. The documentation is sparse, and the service is not fully integrated with the self-service console. Except for the hybrid connectivity and multi-cloud application deployment, not much is known about this new technology from Google.
Here is an attempt to provide a big picture of Anthos.
The Building Blocks
At the heart of Anthos is the most popular open source project of our times – Kubernetes. Anthos is built on the firm foundation of Google Kubernetes Engine (GKE), the managed containers as a service offering on Google Cloud Platform. But other vital technologies augment the power of Kubernetes.
Let’s take a closer look at the core building blocks of Anthos:
1) Google Kubernetes Engine – This is the central command and control center of Anthos. Customers use the GKE control plane to manage the distributed infrastructure running in Google’s cloud, on-premise data center and other cloud platforms.
2) GKE On-prem – Google is delivering a Kubernetes-based software platform that’s consistent with GKE. Customers can deploy this on any compatible hardware and Google will manage the platform. From upgrading the version of Kubernetes to applying the latest patches, Google will treat it as a logical extension of GKE. It’s important to note that GKE On-prem runs as a virtual appliance on top of VMware vSphere 6.5. The support for other hypervisors such as Hyper-V and KVM is in works.
3) Istio – This technology enables federated network management across the platform. Istio acts as the service mesh connecting various components of applications deployed across the data center, GCP, and other clouds. It seamlessly integrates with software-defined networks such as VMware NSX, Cisco ACI, and of course Google’s own Andromeda. Customers with existing investments in network appliances such as F5 can integrate Istio with load balancers and firewalls.
4) Velostrata – Google acquired this cloud migration technology in 2018 to augment it for Kubernetes. Velostrata delivers two significant capabilities – stream on-prem physical/virtual machines to create replicas in GCE instances and convert existing VMs into Kubernetes applications (Pods). This is the industry’s first physical-to-Kubernetes (P2K) migration tool built by Google. This capability is available as Anthos Migrate, which is still in beta.
5) Anthos Config Management – Kubernetes is an extensible and policy-driven platform. Since Anthos’ customers will have to deal with multiple Kubernetes deployments running across a variety of environments, Google attempts to simplify configuration management through Anthos. From deployment artifacts, configuration settings, network policies, secrets and passwords, Anthos Config Management can maintain and apply the configuration to one or more clusters. Think of this technology as a version-controlled, secure, central repository of all things related to policy and configuration.
6) Stackdriver – Stackdriver brings observability to Anthos infrastructure and applications. Customers can track the state of clusters running within Anthos along with the health of applications deployed in each managed cluster. It acts as the centralized monitoring, logging, tracing, and observability platform.
7) GCP Cloud Interconnect – No hybrid cloud platform is complete without high-speed connectivity between the enterprise data center and the cloud infrastructure. Cloud Interconnect can deliver speeds up to 100Gbps while connecting the data center with the cloud. Customers can also use Telco networks offered by Equinix, NTT Communications, Softbank and others for extending their data center to GCP.
8) GCP Marketplace – Google has created a curated list of ISV and open source applications that can run on Kubernetes. Customers can deploy applications such as Cassandra database and GitLab in Anthos with the one-click installer. Eventually, Google may offer a private catalog of apps maintained by internal IT.
Google’s product management team did a great job stacking up the services for Anthos.
Greenfield vs. Brownfield Applications
The core theme of Anthos is application modernization. Google envisages a future where all enterprise applications will run on Kubernetes. To that end, it invested in technologies such as Velostrata that perform in-place upgradation of VMs to containers.
Google built a plug-in for VMware vRealize to convert existing VMs into Kubernetes Pods. Even stateful workloads such as PostgreSQL and MySQL can be migrated and deployed as StatefulSets in Kubernetes.
In typical Google’s style, the company is downplaying the migration of on-prem VMs to cloud VMs. But Velostrata’s original offering was all about VMs. Customers running traditional line-of-business applications such as SAP, Oracle Financials, and Peoplesoft can continue to run them in on-prem VMs or choose to migrate them to Compute Engine VMs. Anthos can technically provide interoperability between VMs and containerized apps running in Kubernetes.
With Anthos, Google wants all your contemporary microservices-based applications (greenfield) in Kubernetes while migrating existing VMs (brownfield) to containers. Applications running in non-x86 architecture and legacy apps will continue to run either in physical or virtual machines.
Broad Industry Partnership
Google is often criticized for the lack of collaboration and partnership with other industry players. With an ambition to beat Microsoft and Amazon, Google is ready to work with established industry players in the ecosystem. This move will undoubtedly help Google in positioning it as an enterprise player.
To succeed with Anthos, Google will have to push its stack through incumbent enterprise players. It is going after the same partners that are hand-in-glove with Microsoft and Amazon to deliver hybrid cloud.
Cisco is extending HyperFlex, ACI, SD-WAN and Stealthwatch Cloud to Anthos to deliver proven software-defined networking capabilities. VMware is adding Anthos support for Pivotal Kubernetes Service (PKS) and SD-WAN by VeloCloud. It is also integrating its popular SDN, NSX, with Istio running within Anthos.
There are over a dozen ISVs supporting Anthos at the launch. From Citrix to NetApp to MongoDB, there are a variety of platform providers integrating their software with Anthos.
Google’s Kubernetes Landgrab
As the founder of Kubernetes, Google knows a thing or two about container management. When Docker started to get traction among developers, Google realized that it’s the best time to unleash Kubernetes in the wild. It also moved fast in offering the industry’s first managed Kubernetes in the public cloud. Though there are over a dozen managed Kubernetes offerings, GKE is still the best platform to run microservices.
With a deep understanding of Kubernetes and the substantial investments it made, Google wants to stake its claim in the brave new world of containers and microservices. The company wants enterprises to leapfrog from VMs to Kubernetes to run their modern applications.
Anthos is a bold move from Google. It is taking a calculated risk in moving away from the clichéd hybrid cloud narrative that its competitors are using to lure enterprises. Anthos is bound to be compared with Microsoft Azure Stack and AWS hybrid story consisting of VMware and Outposts. The fundamental difference between Google and the rest lies in the technology foundation strongly rooted in containers and Kubernetes.
Google wants to capitalize on its pole position in the Kubernetes and cloud native ecosystem. It moved really fast to build an enterprise strategy around Kubernetes. With Anthos, Google aims to become the VMware of container ecosystem. It is precisely doing what VMware did to push its hypervisor and software-defined infrastructure in the enterprise.
But Microsoft is also investing heavily in containers and Kubernetes. It is blurring the line between Azure and Azure Stack by bringing its flagship public cloud services to the private cloud. Microsoft can play back the Anthos track with its Azure Kubernetes Service. It will be interesting to see what Redmond wants to do with Kubernetes on Azure Stack.
If everything plays out in Google’s favor, Anthos will eventually be the preferred platform to run enterprise workloads.
Statement on Enterprise-readiness
Getting the executives from Cisco and VMware on stage with Thomas Kurian was a statement from Google that they are now ready for the enterprise. This certainly increases the confidence of enterprise customers to invest in GCP and Anthos.
Kubernetes is perceived to be a highly technical, geeky platform that appeals to developers and operators. With Anthos, Google is all set to change the perception of Kubernetes. It has turned the platform into a viable, reliable, enterprise-grade hybrid cloud platform.
With Thomas Kurian at the helm, Google has its eye on the enterprise. If it leverages the partnerships and continues to collaborate with the key industry players, Google is all set to shake up the enterprise infrastructure market with Anthos.
Great opportunity for the Cloud Native Ecosystem
Anthos increases the confidence of the cloud native ecosystem. As I mentioned earlier, Google is trying to be the VMware of Kubernetes. But the market dynamics are very different compared to the times when VMware established itself as an enterprise leader.
The most significant difference is the open source software. Google is competing in an environment where software is no more the key differentiator.
To succeed with Anthos, Google has to rely on the community and ecosystem. This opens up avenues for emerging startups delivering niche products that make the cloud native computing stack complete. Startups like Tigera, Portworx, Robin, Confluent, Cloudbees will tremendously benefit from Anthos push.
Service providers and system integrators are gearing up for the Anthos opportunity. From small, local players to global SIs like Accenture and Cognizant, Google’s hybrid push translates to multimillion-dollar services opportunity.
With Google evangelizing and selling Kubernetes to enterprises, many emerging startups will ride the wave to push their agenda to enterprise.
Google’s big bet on Anthos will benefit the industry, open source community and the cloud native ecosystem in accelerating the adoption of Kubernetes.