Public preview: Automatic scaling with Azure Virtual Machine Scale Sets flexible orchestration mode
Published date: September 07, 2021
We are announcing that we have enabled elastic virtual machine profile and automatic scaling for Azure Virtual Machine Scale Sets with flexible orchestration elastic profile and automatic scaling. The features are now in public preview, and provide:
- Up to 1000 instances in a scale set (general purpose virtual machine sizes only)
- Ability to manually add VM instances to the scale set
- The option to spread instances across fault domains automatically, or specify a fault domain
- Place on demand and Spot VMs in the same scale set
- (New) Define a VM profile and specify instance count
- (New) Automatically scale out and scale in based on metrics, schedule, or AI prediction (private preview)
- (New) In guest patching that respects high availability / FD constraints
- (New) Automatic extension updates
- (New) Automatic instance repair/replacement of unhealthy instances
- (New) Terminate notification for on demand and Spot VMs
- (New) Secure by default networking - customers must explicitly define outbound connectivity
- (New) Improved scale out and scale in reliability, latency, and elasticity
This release provides improved quality and simplified Azure customer experience, giving you the opportunity to combine the functionality of Availability Sets and Virtual Machine Scale Sets:
- More easily deploy large scale, highly elastic stateful workloads
- Rapidly scale out and scale in on demand or Spot virtual machines based on metrics or a schedule
- Stateful applications within an availability zone in conjunction with Azure Ultra Disks
- Simplify deployment and management of traditional N-tier applications
Learn more about Virtual Machine Scale Sets with Flexible Orchestration Mode.