Argo Workflow + Spark Operator Or Spark with Knative leads to Server less Deployment

292 Views Asked by At

Can anyone illustrate the difference between deploying the Spark Operator with Argoflow & Apache Spark with Knative. Which route will help to achieve a pure serverless application stack over Kubernretes.

1

There are 1 best solutions below

0
On

I'm not sure what configuration you're imagining for Apache Spark + Knative. Knative is intended to provide developers a serverless deployment and management experience, but AIUI Spark is an independent project which does not specifically target Kubernetes execution.

Similarly, it looks like Argoflow aims to simplify deployment of Kubeflow (ML training) applications specifically, whereas Spark is a general-purpose data processing framework. Your question is a bit like asking "what's better for getting around, a Tesla or a Cessna"? Without knowing where you're going, it's a bit hard to make a recommendation.