This week Microsoft released a patch for their hardware virtualisation platform, Hyper-V, that was highly recommended. It was the first patch of its kind in the 2 years that this technology has been publicly available. It was important enough for us to bypass our normal patch change control process – we take matters of security and stability very seriously.Normally, installing a patch means customer down time. That’s not necessarily so true with our platform.
Our customers running highly available virtual machines on our Hyper-V cluster never missed a beat when our engineers started deploying this update. The virtual machines were simply “Live Migrated” from one physical machine to another. That meant we could operate on the physical hosts without impacting our customers’ machines. The hosts were patched and rebooted. We cycled the virtual machines through the cluster without them dropping a single network packet. This allowed all hosts to be updated within an hour.
Related posts:
