New Project to Help Mesos Container Orchestration Connect Windows and Linux Worlds

By David Hamilton

At MesosCon last week, Microsoft announced the Mesos on Windows Server project, which aims to port Mesos’ highly scalable and elastic container orchestration software directly to Windows Server. This would make it possible to combine both Windows Server and Linux machines into a single pool of resources in their data center or cloud.

Mesos on Windows Server is an open source project and part of the core Apache Mesos, and is still in a technical proof-of-concept stage. Up until now, it has been a collaboration between Microsoft and Mesosphere, but others will be able to join in the effort, and the all code will be freely available on GitHub in the coming weeks.

In April, Microsoft Azure began supporting Mesosphere’s Datacenter Operating System (DCOS), which expanded the options available for container deployment and orchestrations solutions on Azure.

In a Thursday blog post, Microsoft’s Server and Cloud Platform Team wrote, “Mesos is one of the most established and proven open source container orchestration technologies, used in very large scale production environments by several Fortune 500 companies. Furthering our commitment to choice and flexibility for our customers, we have already provided first-class support for Mesos, along with Docker Swarm/Compose, Deis, and many others, on our rapidly growing Microsoft Azure public cloud platform – and today’s announcement is just another step in that journey.”

According to rumors surfacing last week, Microsoft has been interested in buying Mesosphere. News site The Information reported that the companies were in talks regarding an acquisition, and that Mesosphere had its value assessed at just over $1 billion.

Another major announcement from MesosCon was the launch of Mesosphere Infinity, a product for real-time analytics tool that can be used in DCOS. Developed in collaboration between Mesosphere and Cisco, Mesosphere Infinity installs an environment for handling various data-processing tasks and includes a data processing stack consisting of Spark, Akka, Cassandra, Kafka, and Apache Mesos.