In a flow, is it possible to invoke for-loop parallelism on a single worker, to save on cost
Hey!
So I have a flow which uses an SDK which is not very optimized and does a slow API call. This ends up racking up a lot of worker execution seconds, and Windmill cloud bill along with it.
I don't want to ditch the usage of the SDK. One solution would be to re-factor the slow to try to do all the necessary slow calls in a single step, i.e. I would handle the for-loop in the .js level, not the flow level.
I would prefer to keep the flow structure and optimize on the level of Windmill worker usage. Any way to do this?
So I have a flow which uses an SDK which is not very optimized and does a slow API call. This ends up racking up a lot of worker execution seconds, and Windmill cloud bill along with it.
I don't want to ditch the usage of the SDK. One solution would be to re-factor the slow to try to do all the necessary slow calls in a single step, i.e. I would handle the for-loop in the .js level, not the flow level.
I would prefer to keep the flow structure and optimize on the level of Windmill worker usage. Any way to do this?
