In what they call Project Tardigrade, Microsoft is aiming to make Azure servers as durable as tardigrades. Among other things, they want to make the cloud app able to survive a platform failure.
Mark Russinovich, Chief Technical Officer of the Microsoft Azure department, outlined a couple of key changes at the Build 2019 conference that was held last week. A tardigrade, in case you’re not familiar with the species, is a water animal – a moss piglet if you will. Why is this relevant? As it so happens, this animal is one of the most durable ones discovered on planet earth. Not only can it survive in extreme temperatures, but in outer space as well.
According to Russinovich’s words, they want Microsoft Azure servers to be like that. The goal is to avoid having to reboot the virtual machines when things don’t go as planned. The pinnacle of Tardigrade design is to freeze the state of the virtual machines in RAM, then transfer it to a fresh server in the event of trouble. The aim is to offer fault-resistant services.
In technical terms, lightweight virtual machines are used as part of the project. Due to encapsulated external dependencies, it’s possible to migrate them across machines without limits.
As well as that, the company aims to roll out additional availability zones all over the world. According to their claims, Microsoft, in comparison to other providers, offers the greatest number of available cloud regions worldwide.