Define Scale Groups

Application scaling helps you optimize your infrastructure by automatically adjusting the number of compute nodes based on a predefined policy. You can define a scale group (a pool of compute nodes) for any web frontend or backend application and provide consistent performance to your end-users.

Before you Begin

  • To save an image of a Virtual Node, it must be powered down. For this, click on the Power Off under the Actions section.

../_images/Power_off_action_button.png

  • Create and save an image of the application you are looking to scale. The saved image is expected to launch your application on the startup. Use tools like systemD or supervisord for this purpose.

../_images/Saved_image_action_button.png

  • Read concepts guide for understanding how Application Scaling works

Steps to Define Scale Groups

../_images/manage_application_scaling_page.png

  1. Go to My Account

  2. Go to services > Application Scaling

  3. Click on Create Scale Group.

  4. Select a Saved Image that can launch your application at startup

../_images/select_saved_image.png

Note

You can use only C series and M series saved images to create a scale group. Deleting the saved image is not allowed when it is associated with a Scale group. You need to first terminate the associated scale group to delete this saved image.

  1. Select a Compute Plan.

Hint

The compute plan for the scale group can be different from the original plan used to create a saved image.

  1. Click On Create

../_images/select_Scale_group_plan_page.png

Hint

The compute plan for the scale group can be different from the original plan used to create a saved image.

../_images/Create_scale_group_detail_page.png

  1. Enter a short name for your scale group.

  2. Select Min nodes based on the minimum expected workload during off-peak hours.

  3. Select Desired nodes you would like to begin with.

  4. Select Max nodes based on the maximum expected workload during peak hours.

  5. Under the scaling policy, set a threshold for CPU Utilization in terms of percentage(%). Here the CPU Utilization is a node performance metric and the percentage is a threshold value that kicks off the scaling process.

  6. Enter watch periods (integer) and period duration (in seconds) to account for random spikes in target metric (CPU Utilization). Learn more about watch period parameters in Concepts Guide.

  7. Enter a cooldown period in seconds. This duration lets the system stabilize between scaling events. Ideally, this is the time that an application node will take to become effective. So consider the node and your application startup time when choosing this value.

  8. Click Deploy Service to launch scale group nodes with selected policy.

  9. After the scale group is created, you will be redirected to the Manage Scale Group page where you can monitor the launch process with refresh status option.

  10. Select the scale group to check the scale group details where you can monitor and configure scale group parameters.

../_images/scale_group_detail_tab.png

  1. Under the Active Node details tab, you can check scale group node details. Also, you can perform actions like ‘access console’.

../_images/active_node_detail_tab.png

  1. Under the Terminated Node details tab, you can check automatically added/removed scale group node details based on the scaling process.

../_images/terminated_node_detail_tab.png

  1. Also, you can perform actions like ‘access console’ or ‘delete’. When a scale group is in Running state, you can configure a load balancer with dynamic groups.