Schedules

Introduction

The main goal of Cost Analyzer is to provide an effective cost management service that helps users minimize the Azure subscription cost based on resource utilization.

The first step towards cost management is to allow users to optimize the resource cost by scheduling them to deallocate/scale down based on the up and down hours of the business.

Schedules in Cost Analyzer is a feature that enables users to create optimization schedules with resources associated by defining the tiers, throughput values, and the resource state with respect to the Up and Down hours of a Week/Month.

Additionally, it allows users to set up recurrence, enabling the optimization schedule to run automatically at regular intervals.

Supported resource types

Cost Optimization supports the following resource types:

1. App Service Plan

• An app service plan's pricing tier can be scheduled to run at a higher tier during Up hours and a lower tier during Down hours.

• The Up and Down config values determine the tiers that should be set to the resources in the Up and Down hours, respectively.

2. Cosmos SQL Container

• The throughput value of a Cosmos SQL Container can be scheduled to run at a higher value during Up hours and a lower value during Down hours.

• The Up and Down config values determine the throughput values that should be set to the resources in the Up and Down hours, respectively.

3. Cosmos SQL Database

• The throughput value of a Cosmos SQL Database can be scheduled to run at a higher value during Up hours and a lower value during Down hours.

• The Up and Down config values determine the throughput values that should be set to the resources in the Up and Down hours, respectively.

4. Cosmos Table

• The throughput value of a Cosmos Table can be scheduled to run at a higher value during Up hours and a lower value during Down hours.

• The Up and Down config values determine the throughput values that should be set to the resources in the Up and Down hours, respectively.

5. Data Factory Pipelines

• Pipeline triggers can be scheduled to start in Up hours and stop in Down hours.

6. Logic Apps

• Logic Apps can be scheduled to run in the Up hours and stopped in the Down hours.

7. SQL Database

• The pricing tier of an SQL DB can be scheduled to run at a higher tier during Up hours and a lower tier during Down hours.

• The Up and Down config determines the pricing tier that should be set to the resources in the Up and Down hours, respectively.

• The DTU count and Data max size can be customized for the DTU-based purchasing model. In contrast, the hardware configuration for Vcore-based purchasing models can be customized.

8. SQL Elastic Pool

• The pricing tier of an SQL Elastic pool can be scheduled to run at higher tier during Up hours and a lower tier during Down hours.

• The Up and Down config determines the pricing tier that should be set to the resources in the Up and Down hours, respectively.

• The DTU count and Data max size can be customized for the DTU-based purchasing model. In contrast, the hardware configuration for Vcore-based purchasing models can be customized.

9. SQL Managed Instances

• The Up and Down config of an SQL Managed Instance can be scheduled to run at a higher tier during Up hours and a lower tier during Down hours (or) initiate a startup/shutdown.

• The Up and Down config determines the pricing tier that should be set to the resources in the Up and Down hours, respectively.

• The Vcore, Data max size, and Hardware configuration can be customized based on the Service tier.

10. Synapse Pipelines

• Pipeline triggers can be scheduled to start in Up hours and stop in Down hours.
  

11. Virtual Machine

• The Up and Down hours can be scheduled to upgrade/downgrade the virtual machine service tier (or) initiate startup/shutdown.

12. Virtual Machine Scale Set

• The Up and Down hours can be scheduled to upgrade/downgrade the service tier of virtual machine scalesets (or) initiate startup/shutdown.

Schedule recommendations

• Cost Analyzer offers schedule recommendations by analyzing resource usage patterns within the cost management group. It suggests users to create schedules with predefined up and down hours to achieve significant savings by shutting down or resizing the resources during non-active periods.

• Schedule recommendations summary can be accessed by navigating to Optimization -> Schedules within a cost management group.

How to setup recommendations?

Follow the below steps to receive schedule recommendations for a Cost Management group:

1. Go to Recommendation settings and enable the Schedule Recommendations toggle in General Settings
2. Define values for currency, frequency and timezone to receive customized recommendations.

3. Configure metrics and threshold values defining the downtime of the resources within this Cost management group.

4. Click Apply.

Clicking the View recommendations option in the Schedule summary allows you to view recommended resources for scheduling. The widget provides a list of all possible schedules that can be created for the associated resources in the Cost Management group.

Resources eligible for savings are recommended with pre-defined up and down hours, determined by analyzing the usage patterns of the resource’s default metrics.

Clicking on the resource name displays the cost details, along with potential savings, recommendations to scale down to a smaller size or shut down during off-peak hours, and suggested scheduling hours. This helps reduce resource costs during periods of inactivity.

Create schedules based on recommendations

1. Click on the name of the recommended resource in the schedule recommendations blade
2. Click the Add to schedule option in the recommendation details blade.
3. Provide basic schedule details and click Next.
4. The scheduling hours configuration is defined based on recommendation and can be adjusted. Additionally, recurrence can be set to run the schedule at regular intervals.
5. Only the resource recommended for the schedule will be available to configure in the Resource Selection screen.
6. Choose the desired notification channels and create the schedule.

The illustration shown below depicts creating a schedule based on recommendation:

Recommendation Settings

The recommendation settings are split into two categories: General and Resources.

General

The following settings are required to be configured to receive recommendations based on user preferences:

i) The currency in which the recommendation details are displayed.

ii) The time zone for processing the schedule.

ii) The evaluation period for analyzing resource usage, which is considered in providing the recommendation.

Resources

• Schedule recommendations are based on default metrics like CPU and memory consumption, as well as other standard metrics. Users who need to include additional metrics for their resource recommendations can customize the metric configuration.

• For resource types like Virtual Machines, Virtual Machine Scale Sets, and SQL Managed Instances, users can choose to either stop the resources during off-peak hours or downsize them.

• Enabling the Family restriction checkbox for Virtual Machine (or) Virtual Machine Scale sets ensures that their schedule recommendations are within the bounds of the Current SKU family. This option can be used when you want to exclude certain non-production VM families from being recommended for your production VMs.

• Excluding VM families for Virtual Machines or Virtual Machine Scale Sets ensures that their recommended SKU do not come from the excluded VM families.

Recommendation refresh

• Schedule recommendations for a cost group will be updated every two days based on the last updated date range.

• It is also possible to perform a manual refresh to view recommendations instantly.