continueChat with specific blockId
My typebot flux has a lot of HTTP requests and sometimes a request takes too long to resolve and my API request to continueChat timeout, is there a way to retry the request to this specific stage/blockId and guarantee idempotency, and be able to continue the flux after the retry on this specific portion?
5 Replies
The problem is not really increasing the timeout as I do these request from a serverless environment and not is the best to keep increasing the timeout.
What I would like is this idempotency to retry my request somehow, maybe when calling the
continueChat endpoint I receive the next blockId and being able to request this specific blockId then i would be able to retry this request.We decide to controle everything internally.
What would be your retry mechanism? Maybe that's what we should offer as an option directly on Typebot
Im using aws lambda to handle this process of sending message to typebot and etc, if i could retry i would send this http request
continueChat to a sqs queue and in case of any error just retry the queue
or with we could pass a parameter to move the cursor to the previous state like HEAD~1Thanks for the feedback