Azure customers use Service Fabric to manage both stateless and stateful microservices at scale. While our customers love the reliability, scalability, and richness of Service Fabric’s features, they have asked us to simplify the service experience. Azure Service Fabric managed clusters remove complexity associated Service Fabric clusters by simplifying deployment and management operations.
Today we are announcing the general availability (GA) of several popular features focused on enhancing networking, node types, and scaling capabilities with API release 2022-01-01. Customers such as Power Query Online, Azure Backup, Hardware Health Service, and Azure DNS have requested and leveraged the features for their workloads. During this period, their feedback and validation has created further confidence in these new and existing capabilities of Service Fabric managed clusters. With this GA release we continue to demonstrate our commitment to making Azure Service Fabric managed clusters the ideal choice for Azure Service Fabric customers. All GA features and APIs are covered by the standard Azure service level agreement (SLA) and can be used in production workloads.
New GA Features
Network enhancements:
Bring your own Virtual Network
Enables customers to allocate a dedicated subnet in an existing virtual network the managed cluster will use for IP allocation needs. See bring your own virtual network for more information.
Bring your own Load Balancer
Bring your own load balancer allows you to use an existing pre-configured Azure Load Balancer for secondary node types for both inbound and outbound traffic. We also enable customers to configure an internal-only load balancer by using the default load balancer for external traffic. See bring your own load balancer for configuration information.
Managed Load balancer improvements
Customers can now easily configure any available load distribution policy on the managed load balancer. This allows customers more flexibility in scenarios that the managed load balancer can support such as session persistence. See how to configure load balancer policy for your cluster.
IPv4/IPv6 dual stack support
Enables IPv4/IPv6 dual stack for the managed resources such as the Load Balancer and VMs. See enable IPv6 to configure for your cluster.
Auxiliary Subnets
Auxiliary Subnets provide the ability to support additional network scenarios that do not require a node type such as private link service and bastion hosts. Customers can now easily add additional Subnets for scenarios like these on the managed cluster resource. See configure network settings for more information.
Accelerated Networking
Accelerated networking enables single root I/O virtualization (SR-IOV) to a VMSS VM that is the underlying resource for node types. This high-performance path bypasses the host from the data path, which reduces latency, jitter, and CPU utilization for the most demanding network workloads on supported VM types. See the node type configuration options for how to enable accelerated networking.
Node type enhancements:
Host Level Encryption
Encryption at host improves on Azure Disk encryption by supporting any OS type and image, including custom images, for your VMs by encrypting data in the Azure Storage service. This method does not use your VM's CPU enabling workloads to fully utilize all the VM SKUs resources. See enable disk encryption for how to configure on a node type.
Multiple Managed Disks
Customers can now configure additional managed disks to isolate workloads from Service Fabric runtime data disk. Additionally, you can configure a specific drive letter associated for each additional managed disk. See node type configuration options for how to configure additional disks.
Configure Primary Data Disk Drive Letter
Customers can now specify the drive letter for the primary data disk. See node type configuration options for how to configure.
Stateless node type temp disk support
Stateless node types can now be configured to utilize the ephemeral temporary disk as the data disk instead of a managed disk. Using a temporary disk can reduce costs for a cluster. See stateless node types for how to configure.
Scaling and lifecycle management:
Autoscaling
The managed cluster model now includes an integrated autoscaling solution. The integrated solution provides 3-6x faster end to end scaling operations while being safer and more predictable. See the autoscaling how-to to configure for your workloads.
Automatic Extension upgrades
Managed clusters now support automatic extension upgrades for supported extensions. This reduces operational overhead and keeps extensions updated. See how to enable for more information.
Learn more
If you would like to learn more about the service in general, head over to the Service Fabric managed clusters documentation or use the following Quickstart. We also have additional resources including how-to documentation and cluster templates. You can view the public feature roadmap and provide feedback on the Service Fabric GitHub repo.
