Azure is rebooting your server. Nobody notices.

Frequently asked questions

Why can't a single VM survive Azure's own maintenance?

Azure periodically reboots the physical host underneath a VM for planned maintenance — a single VM has no way to stay up through that, no matter how well the application itself is built.

What does an availability set actually do?

It spreads your VMs across separate fault domains (different racks and power) and update domains (different maintenance groups), so planned maintenance or a hardware fault takes out one VM at a time while the load balancer's health probe keeps sending traffic to the ones still up.

Read deeper on Azure

Twenty seconds gets the shape of an idea across. These go into how it behaves in production.

• AI Bot Attacks on a Mobile App Login, and How Azure Stops Them

  A script tries 8,000 passwords a minute against your login API at 3am. Rate limiting, smart lockout and Defender for Cloud make speed the losing move.

• Azure Key Vault Explained: The One Secret Tutorials Never Remove

  Moving every secret into Key Vault leaves exactly one behind: the credential that opens the vault. A managed identity is what removes it.

• How to Secure a Mobile API on Azure in Five Steps, in the Right Order

  HTTPS only, authentication, authorization, secrets and monitoring: five steps to secure a mobile API on Azure, and why the order carries most of the value.