Ultra Task FAQs

What do Ultra Tasks do?

Provide the ability to call pipelines as an API with low latency response times (milliseconds).

What are the requirements for setting up Ultra Tasks?

An Ultra Task should run on a dedicated Snaplex. If the pipeline executed by an Ultra Task has one unconnected input view and one or more unconnected output views, then it requires a FeedMaster.

Do Ultra Tasks support all Snaps?

Not all Snaps can be used in a pipeline run by an Ultra Task. Snaps with limitations can be used in a child pipeline, which you can invoke locally using the Pipeline Execute Snap.

How do I achieve High Availability?

Add multiple FeedMasters, Load Balancers, and Virtual IP.

Is there a node restriction for Ultra Tasks?

Generally, no. However, there are concurrency best practices for determining the proper number of nodes.

If my Tasks are internet facing, can I use other types of authentication?

Yes. The bearer token is the default authentication method. For more fine-grained access control, you can implement JWT authentication in the pipeline using JWT Snaps. JWT allows for multiple tokens per task and supports access control at the project level or based on other criteria. You can also disable the bearer token and use API Policy Manager for additional authentication and authorization policies.