Responding to customer hunger for bigger virtual machines, Microsoft has bulked up its most powerful VM with more storage and higher throughput, paving the way for running large-scale, database-driven applications in the Microsoft Azure cloud.
The newly launched GS-Series is based on the G-series of VMs, which proved to be surprisingly popular since they were launched in January, said Corey Sanders, Microsoft Azure director program management. Usage of this VM has doubled by 50 percent in the last three months, he said.
The new VMs are not currently available from Azure's Australian region; the can only be spun up in the West US, East US 2, and West Europe regions.
The G series, which runs on the Intel Xeon E5 v3 processors, is the most powerful VM offered on Azure. Customers use the G-series to run databases -- including Microsoft SQL Server and MySQL, and even NoSQL databases such as MongoDB -- that power internal and public-facing applications, as well as serve as the cornerstone for data warehouse-style data analysis, Sanders said.
Customers were also lashing together hundreds of the G-series VMs to run them as clusters to power large applications.
"The additional throughput of the GS-series will make it even easier to run these types of workloads on Azure," Sanders said.
The GS-series is designed to ease the process of deploying these large workloads by including large amounts of storage and by upping the IOPs (input/output per second).
The GS-series offers up to 64 terabytes of storage, provides 80,000 IOPs and can deliver 2,000 MB/s of throughput to the backend storage.
Both the G and GS-series VMs will now offer 20 Gbps (gigabits per second) of network bandwidth.
In addition to the new VMs, Azure now also features a new diagnostic tool that allows administrators to see the serial and console output from a running VM, which can help in troubleshooting.
Microsoft also introduced a new service bus for exchange data across different VMs, called the Azure Service Bus Premium Messaging, which is ideally suited for systems running on micro-service architectures with lots of individual components.