When running a long task, I would like to be able to pick it up if a worker gets lost, but it means I have to set an extremley long timeout to allow for tasks of varrying length to increase.
Ideally what I would like to do is reclaim a message during a long process to extend the timeout for tasks that I know are going to be long running. The problem is, right now the only task_id we have is the TaskIQ id in the data, not the stream id. Is it possible that when we receive a task from the broker, we can get the stream task_id from redis and store it in either state, or wherever else might make sense.
When running a long task, I would like to be able to pick it up if a worker gets lost, but it means I have to set an extremley long timeout to allow for tasks of varrying length to increase.
Ideally what I would like to do is reclaim a message during a long process to extend the timeout for tasks that I know are going to be long running. The problem is, right now the only task_id we have is the TaskIQ id in the data, not the stream id. Is it possible that when we receive a task from the broker, we can get the stream task_id from redis and store it in either state, or wherever else might make sense.