Policy Configuration

A policy handles snapshot and backup scheduling for specific workloads. It is important not to make a policy cover too many workloads or too few in order to make scheduling and processing simple.

This section covers supported policy maximums. Numbers might vary depending on deployment types and available resources. These values represent what is currently supported, also in light of acceptable performance.

The tables in this section use the number of objects and their variations as a metric for estimations.

Recommended Configuration

These values are based on Veeam QA testing and the recommended maximums per policy for the specified configuration.

When protecting 50 workloads in a policy you can expect around 7.3 GB/s writing speed (using 50 workers with a VM size of F8s_v2) which falls under the ingress of Azure regions (about 7.5 GB/s or ~ 60 Gbps). For more information about Microsoft Azure regions ingress limits, see the Scalability and performance targets for standard storage accounts documentation.

It is possible to protect more then 50 workloads in a policy. This does require configuring the load options for the target repositories by using the “Limit concurrent backup task” setting.

Maximizing Throughput

The amount of simultaneous workers that can process a policy is determined by the speed at which you can write to the backup repository storage account. This requires some consideration as you want to maximize policy processing throughput.

For example, one backup policy can only write to one storage account. When using a worker size of F2s_v2 with 80MiB/s throughput to a storage account that can handle 25 Gbps, this can have a maximum of 3GB/s of throughput to the storage account, so a maximum of 38 workers. This means that the policy, while it can contain around 50 workloads to protect, should only have 38 workers processing simultaneously.

As mentioned before, if required, you can configure a limit of concurrent backup tasks in the repository settings.

---