Deploying your container images on a serverless platform means less resource management and maintenance for you or your business. But which platform is the best? We explore Google Cloud Run and Microsoft Azure Container Instances (ACI) and discuss which cloud providers perform well with Docker images and which is the best value.
Achieve Higher productivity with Iron
Need a Worker service that’s completely scalable? Talk to Iron.io today about a free trial of IronWorker.
Table of Contents
Google Cloud Run Basics
GoogleCloud’s Google Cloud Run is a serverless platform for deploying containerized applications. It supports a variety of languages including:
A Cloud Run Service is a container managing a set of routes and configurations. Use cases for these services include defining resource ownership and rollout policies.
Google, Knative, and Kubernetes
Cloud Run builds on Knative. Knative is an open source extension of Kubernetes, a tool used for managing container clusters. Knative disguises some of the complexities of the Kubernetes engine, simplifying container management and empowering developers to concentrate on coding.
Microsoft Azure Container Instances Basics
ACI is another server-less cloud platform for running containers. Their main USP is that users don’t need to manage virtual machines, plus ACI provides hypervisor isolation. This means containers always run in isolation without having to share a Linux kernel.
Azure Kubernetes Service – AKS
AKS is Azure’s way of managing Kubernetes clusters. Clusters are groups of containers deployed at once to deal with spikes in traffic. Google’s equivalent is GKE clusters. Autoscaling and load balancing help provide additional compute power just when it’s required.
Shouldn’t I Be Using AWS Fargate, AWS Lambda, or Similar from Amazon?
Both AWS Fargate and Lambda support the deployment of container images, but AWS isn’t the only provider out there. Despite their global dominance in many cloud native compute solutions, AWS has limitations. Also, pricing varies between services. For example, Fargate charges per vCPU while Lambda charges per execution of function.
If you’re already using AWS services like ECS or EKS, you may naturally gravitate toward other Amazon web services. However, it’s great to know that there are plenty of other options out there — many focused directly around Docker containers and their functionality, just like IronWorker.
Iron.io Serverless Tools
Find out more about IronWorker and how it provides a scalable and easy-to-use way to manage your tasks asynchronously.
Google Cloud Run vs. ACI: Pricing
Anyone can try Cloud Run for free. New users get $300 to spend on the service for 90 days, and all users get 2 million requests free. Epitomizing the phrase, “The first one’s free, but you’ll be back,” Google is confident that developers will want more from Cloud Run than the free offerings. However, for a startup, Cloud Run could be a completely free option for a limited time.
ACI boasts no upfront cost, but beyond trying for free, businesses have to request a quote. Pricing is based on CPU seconds consumed.
Google Cloud Run vs. ACI: Which Works Best with Docker Containers?
Google Cloud run works as a PaaS, creating publicly accessible URLs for any containers deployed. Cloud Run deploys within seconds and is very easy to use for most developers. It currently does not support Kubernetes pod.
ACI was the first serverless container platform in the public cloud. Hence, it’s pretty agile. Some limitations include not being able to accept existing Kubernetes pod definitions due to using its own specs, but it can mimic a pod definition, unlike Cloud Run. ACI is also persistent by default.
The Bottom Line for Serverless Container Orchestration
The main functions you need from your container deployment platform are:
· Scalability: A scalable platform means you use and pay for only the resources you need in the understanding that resource requirements will dip and peak depending on the tasks at hand.
· Ease of use: Even the most hardcore DevOps team likes shortcuts as long as there’s no loss of functionality.
· Performance and speed: Deployment should take seconds and performance should impress, not depress.
IronWorker is a fully hosted, background job processing tool designed from the ground up to work with Docker and containers.
Unlock your full potential with Iron.io
Whether you need to process your ETL, automate email delivery, or optimize your mobile compute solutions, IronWorker is a scalable solution. Get in touch to find out how to get started.