Scaling Service Management Automation

Important

This version of Service Management Automation (SMA) has reached the end of support. We recommend you to upgrade to SMA 2022.

Use the guidance in this section to scale out a machine tier in a service that's deployed in Service Management Automation (SMA). You can add runbook workers and web services to add additional capacity for runbook processing.

Initial recommendations

The recommended configuration is as follows:

  • Three VMs, each with an installed runbook worker and web service.
  • The incoming web traffic should be load balanced.
  • The machines should have at least two cores and contain a minimum of 4 GB of RAM, along with 60 GB of storage.
  • Only one PowerShell module should be installed.

SQL Server recommendations

SQL Server recommendations include:

  • For the SQL Server database, 8 GB of RAM and 8 cores are recommended.
  • One month of data under heavy load (12 jobs per minute for a month) results in 20 GB of database space usage. By default, job purging should keep the space usage from growing much beyond this. Learn more.

Scale out

Scale out recommendations:

  • If runbook jobs are running slowly, you can increase the number of runbook workers that are sharing workloads. New runbook worker/web service instances must be installed on their own VMs. Before installing or uninstalling a runbook worker, stop the Runbook Worker service (rbsvc) on the computer on which the runbook worker is installed. Learn about runbook worker deployments.