Microsoft has removed fees for transferring data between different availability zones within the same cloud region.

“We are announcing that Azure will not charge for the data transfer across availability zones regardless of using private or public IPs on your Azure resources,” the company said this week. “With this change, Azure will further encourage and support customers’ efforts in building more resilient and efficient applications and solutions on Azure.”

Availability zones are separate groups of independent data centers within a single cloud region. The zones are far enough apart they are likely to avoid large single events that could affect uptime while offering inter-zone latency of around 2ms.

Previously, data transfer within the same Azure availability zone was free, but transfers between two different availability zones cost $0.01 per GB. Data transfers from an outside environment into Azure have always been free.

Microsoft pitched the change as a way to increase and encourage resiliency and redundancy, making it cheaper for customers to deploy applications across multiple availability zones.

Microsoft still charges for data transfers between Azure regions. Costs vary, starting at $0.02 per GB between regions in the US or within Europe, and up to $0.16 per GB between regions in South America.

Inter-continental data transfer between Azure regions starts at $0.05 per GB and reaches up to $0.16 per GB, depending on which region the data is being transferred from.

For Amazon Web Services, data transfer within the same availability zone is also free. However, depending on the services being used, Amazon often charges to move data between availability zones within the same region, and between local zones and availability zones within the same region. It also charges for inter-region data transfers.

Google also often charges for data transfer between different availability zones of the same Google Cloud region and between different Google Cloud regions, again depending on the services being used.

Amid pressure from regulators, Microsoft, Google, and Amazon are all ending egress fees for customers looking to exit the cloud entirely. However, most still charge for companies looking to routinely move data between different cloud providers.

Microsoft launches preview of Compute Fleet service

This week also saw Microsoft launch a new large-scale virtual machine (VM) provisioning service in preview.

The company said Compute Fleet is a new service that simplifies the provisioning of compute capacity across different VM types, availability zones, and pricing models in a single API call to “help customers achieve the scale, performance, and cost they require.”

“Embracing the 'Fire & Forget-it' model, we are introducing Compute Fleet to automate your deployment, management, and monitoring of virtual machines (VMs) without requiring intricate code frameworks,” Microsoft's Azure principal product manager Rajeesh Ramachandran, said in a blog announcement. “Tell us what you need, capacity- and instance-wise for standard and Spot VM, and Compute Fleet will create both standard and spot VMs from a customized SKU list, tailored to your workload requirements with the capability to maintain the spot VM target capacity in the fleet.”

The service can reportedly deploy up to 10,000 VMs across multiple SKUs and sizes. It is currently available in the East US, East US2, West US, and West US2 regions.

Microsoft said the service is useful for large enterprises needing to deploy and oversee thousands of VMs concurrently; cloud-native organizations requiring flexibility in deploying and managing latency-sensitive workloads; and companies aiming to amalgamate pricing models and VM types within a single application.