Asset Compute projects are Node.js projects, generated using the Adobe I/O CLI, that adhere to a certain structure that allow them to be deployed to Adobe I/O Runtime and integrated with AEM as a Cloud Service. A single Asset Compute project can contain one or more Asset Compute workers, with each having a discrete HTTP end-point referenceable from an AEM as a Cloud Service Processing Profile.
Click-through of generating an Asset Compute project (No audio)
Use the Adobe I/O CLI Asset Compute plug-in to generate a new, empty Asset Compute project.
aio app initto begin the interactive project generation CLI.
WKND AEM Asset Compute
Actions: Deploy runtime actions
DX Asset Compute Worker v1
The developer tool requires a file named
console.json that contains the necessary credentials to connect to Adobe I/O. This file is downloaded from the Adobe I/O console.
console.jsoncredentials for, in this case select
.jsonfile prefixed with the project and workspace, for example:
Rename the file as
console.json and move it in the root of your Asset Compute worker project. This is the approach in this tutorial.
Move it into an arbitrary folder AND reference that folder from your
.env file with a configuration entry
ASSET_COMPUTE_INTEGRATION_FILE_PATH. The file path can be absolute or relative to the root of your project. For example:
The file contains credentials. If you store the file within your project, make sure to add it to your
.gitignorefile to prevent from being shared. The same applies to the
.envfile – These credentials files must not be shared, or stored in Git.
The final Asset Compute project is available on GitHub at:
GitHub contains the final state of the project, fully populated with the worker and test cases, but does not contain any credentials, that is,