Pin applications to engine sizes

Overview

Qlik Cloud automatically places applications on compute engines based on calculated metrics that determine the required resources. This automatic placement works well for most apps, but some require manual intervention.

Apps with large chart objects that generate large hypercubes can consume all available memory and crash. By pinning an app to a specific engine size, you control the minimum compute resources available for both reloads and interactive consumption.

If you’re new to this topic, first read the managing analytics applications overview.

Prerequisites

• The app must be in a space with Large Apps support enabled.
• An API key for a user with the Tenant Admin role.
• The app ID and tenant URL.

Understanding engine placement

When you pin an app, it runs on large app engines and consumes large app capacity. You can track this usage through quota events.

Valid engine sizes

Determining which engine size to use

You should select an engine of an appropriate size relative to your application. Qlik by default places applications on engines four times larger than the size of the application in memory.

For example:

• For a 10 GB application, it would be 10 * 4 = 40 and use a 40 GB pod.
• For a 25 GB application, it would be 25 * 4 = 100 and use a 120 GB pod, the next available size.

When pinning apps, you are forcing a minimum size engine for your app. Qlik will use the minimum, or a larger size engine if required by the app.

Quota consumption

Pinned apps consume quota from your Large Apps capacity. When you pin an app, it will use the greater of:

• The pinned capacity divided by four, for example:
  • On a 40 GB engine, 40 / 4 = 10, therefore 10 GB.
  • On a 120 GB engine, 120 / 4 = 30, therefore 30 GB.
• The actual size of the application in memory (visible in the administration console, or on the app details page).

Set an engine size override

To pin an app to a specific engine size, call the placement endpoint with the minEngineSize attribute in the payload:

curl -L -X PUT "https://<TENANT>/api/v1/apps/<APP_ID>/placement" ^
-H "Authorization: Bearer <ACCESS_TOKEN>" ^
-H "Content-Type: application/json" ^
-d "{
    \"minEngineSize\": \"40\"
}"

Replace <TENANT> with your tenant hostname, <APP_ID> with the app GUID, and <ACCESS_TOKEN> with your API key or OAuth access token.

A 200 response indicates success. Omitting the minEngineSize attribute sets the value to 0, which removes any existing override.

Verify the override

To retrieve the current placement override for an app:

curl "https://<TENANT>/api/v1/apps/<APP_ID>/placement" ^
  -H "Authorization: Bearer <ACCESS_TOKEN>"

The API returns 200 if an override exists, or 404 if no override exists.

Remove an engine size override

To delete the override and return to automatic placement:

curl -X DELETE "https://<TENANT>/api/v1/apps/<APP_ID>/placement" ^
  -H "Authorization: Bearer <ACCESS_TOKEN>"

The API returns 200 if the override was removed, or 404 if no override existed.

Monitor capacity consumption

Pinned apps consume large app capacity. The quota events track:

• Quota consumed when an app opens or reloads
• Quota released when an app unloads
• Allocation failures when capacity is insufficient

These events are available via the Audits API for tenants with Large Apps support.