Overview of Operations

To use the Akridata System, users interact with the AkriManager console to define, deploy, and operate data pipelines. Programmatic support for all AkriManager operations is also possible, using REST API calls.

The AkriManager service supports multiple user personas — admins, who are responsible for setting up the microsites and data sinks, and onboarding other users; users, who specify and access the results of data pipelines; and operators, who participate in the operation of a data pipeline.

Data pipelines are specified using the workflow concept, which captures the processing, prioritization, transfer, and storage of source data into the target data sinks. AkriManager provides the following functionality:

• compose workflows^[1] from pre-existing processing modules,
• browse and retrieve the meta-data and data objects resulting from workflow execution, and
• monitor workflow instances, and microsite and data sink utilization.

Each workflow runs within an Akrinode. For the purpose of scaling and orchestrating multiple workflows across multiple Akrinode clusters, Akrimanager provides a construct called workflow groups. 

^[1] The capability to compose workflows will be released in a future release.