Google announced a number of new products at the recently concluded Cloud Next event in San Fransisco. Categories that saw new product announcements range from Serverless and Data Management to Networking and Security. However, the one that made the biggest splash is in the hybrid and multi-cloud category; Google Anthos.
Anthos refers to Google’s portfolio of tools that enable organizations to move towards a hybrid and multi-cloud architecture. With Anthos, customers can deploy, run and manage their applications on-premise, in the Google cloud and on other cloud providers including AWS and Azure.
Almost all public cloud providers have some sort of a hybrid cloud offering. This is not surprising given the fact that only 10% of the total workloads running on-premise have moved to the cloud. Out of these nearly 80% have moved back. Given these numbers, cloud providers have all the more reason to bulk up their hybrid cloud offerings.
In addition, there has been a growing clamour for a multi-cloud deployment and management platform. High availability, reliability, and the freedom to pick and choose products and services from the different cloud providers are at the core of this movement. Fear of vendor-lock is a very real thing for today’s enterprise. With its support for Azure and AWS, Anthos tries to bridge this gap.
GKE, GKE on-prem, Istio on GKE, Anthos Migrate (beta) and Anthos Config Management are the core components of the Anthos portfolio of tools. It also has built-in support for platform level services including Stackdriver, Cloud Build and Binary Authorization.
Let’s take a quick look at each.
GKE is the managed Kubernetes offering of Google. GKE makes it easier to deploy, manage and scale containerized applications on the Google cloud. The environment GKE provisions consists of multiple compute engine instances pooled together to form a cluster. Under the umbrella of Anthos, GKE will extend this environment to other cloud providers including AWS and Azure as well as on-premise.
GKE on-prem brings all the advantages of the Google Cloud ecosystem to on-premise data centres. With GKE on-prem Kubernetes clusters running both on Google’s own infrastructure as well as on-premise can be managed under a single pane of glass view.
Anthos Migrate is a cloud migration tool which allows applications running on VMs from other cloud providers or on-premise to be containerized and moved to GKE. Anthos Migrate does all this while maintaining the application state and with minimal downtime.
Istio provides security, connectivity and monitoring for microservices. With Istio microservices distributed across cloud providers and on-premise infrastructure can be connected securely. In addition, Istio also provides tracing, monitoring and logging for the service mesh.
Anthos Config Management allows the creation of uniform configurations reflecting administrative policies in one central tool. These configurations can then be applied to Kubernetes clusters running across cloud providers and on-premise. Configuration policies can range from RBAC and resource quotas to network and admission control policies.
Both AWS and Azure have dabbled in the hybrid cloud space with AWS Outposts and Azure Stack.
AWS outposts has at its core the same operating idea as Anthos; provide a consistent developer experience across on-premise and the AWS cloud. For this to happen AWS delivers fully managed compute, storage and networking equipment to customer locations. This equipment is then used by AWS Outposts to create a seamless hybrid cloud with access to the same AWS services that are available in native AWS data centres. The VMware variant of AWS Outposts comes with VMware’s SDDC stack loaded on top of the hardware provided by AWS.
Azure Stack is Azure’s hybrid cloud solution allowing enterprises to bring native Azure services to a private on-premise data centre. Initially billed as a solution it later pivoted to a partnership with hardware vendors and required their hardware to be installed before Azure Stack could be used on-premise. There are two variants of Azure Stack; a managed service where the service provider manages the hardware and software and one where the hardware is bought and hosted locally with Azure Stack running on top.
As is apparent, both AWS Outposts and Azure Stack are restricted to on-premise and the respective cloud provider itself, with no support for other cloud providers in the pipeline as of now. With Azure Stack’s pivot, it joined AWS Outposts in requiring both the hardware and software to be installed locally before it can be used on-premise.
With Anthos, Google has tried to reach for the promised land of write once, run anywhere. The introductory post from Google phrases it as “a new open platform that lets you run an app anywhere—simply, flexibly and securely… Anthos lets you run your applications, unmodified, on existing on-prem hardware investments or in the public cloud”.
The operative phrase here is the public cloud. In fact Anthos specifically mentions AWS and Azure as the platforms Anthos supports for workload deployment and management: “Anthos will also let you manage workloads running on third-party clouds like AWS and Azure, giving you the freedom to deploy, run and manage your applications on the cloud of your choice, without requiring administrators and developers to learn different environments and APIs.”
Admittedly other cloud providers including Alibaba, IBM and Oracle are not supported yet. Another important difference with AWS Outposts and Azure stack is the fact that Google Anthos is completely software-based and requires no additional hardware.
To get started with Anthos you first need to reach out to a Google representative to get access. Once given access you then need to download and install GKE on-prem, set up a GKE on-prem cluster, spin up a containerized application in that cluster or migrate one to it using Anthos Migrate and finally register the cluster with GCP.
Once registered you will be able to view and manage all GKE clusters whether on-prem or in the cloud. Next, you can optionally install Istio to secure, monitor and connect any microservices you might have running in the GKE cluster as well as enable Anthos Config Management to consistently manage policies.
Anthos is a monthly subscription service with a minimum one-year commitment and a base price of $10K/month for a 100 vCPU block.
Further vCPUs can be purchased in blocks of 100vCPUs when usage exceeds the preceding block. For example, a workload using 150 vCPUs would require 2 blocks of 100 vCPUs each.
Anthos will also automatically add additional blocks when usage exceeds the amount purchased.
With Anthos Google is looking to beef up it’s hybrid cloud offering and make it more appealing to enterprises. It has also one-upped AWS and Azure by building in multi-cloud support. With Anthos Google has put the customer front and centre, since that’s where the demand for a multi-cloud solution was most persistent.
The fact that Anthos is a software-only solution also makes it more appealing since enterprises no longer have to go through the hassle of updating their on-premise hardware. However, it does remain to be seen how the software-only approach works considering the fact that Azure had to backtrack on it with its hybrid cloud solution.
Getting Ready for Kubernetes in Production? Download the Complete Production Readiness Checklist with Checks, Recipes and Best Practices for Availability, Security, Scalability, Resource Management and Monitoring
Fan of all things cloud, containers and micro-services!
There has been a major uptake in on-premises Kubernetes deployments as of late. In this article we will outline some of the challenges and roadblocks that enterprises face when deploying and operating Kubernetes on-premises. We will also review the replex solution, see how it works on-premises , outline some of its features and how it helps mitigate those challenges.
January 6, 2020
5 min read
A step by step walkthrough of deploying a highly available, reliable and resilient Kubernetes cluster leveraging AWS EC2 spot instances as worker nodes using both Kops and EKS.
October 9, 2019
5 min read
FinOps is a cross domain discipline that represents a set of tools, best practices and processes aimed towards making software and infrastructure more cost effective. In this article we provide an introduction to Kubernetes Finops.
September 10, 2019
5 min read