Another question? for our flyte pipelines 1. to pull the image for every step / task, 2. execute the code and 3 tear down the pod takes time. We have a slow artifactory and the number of pipelines / users is growing which creates more traffic to the artifactory. So can we say to flyte to use the same pod for all the tasks. The pod that has pulled and executing the default image and upload to the pod the relevant code for its step / task.

There are solutions, you can prime the image on all nodes

Thanks! @Ketan (kumare3) I can ask about been open with the usage of flyte. Could you please elaborate more on the possible solutions.

The reason I said was as we talk at the community sync we can discuss ways to solve your problems, would love to understand nature of usage etc