LinkedIn
Twitter
I am creating a Report process that once a month generates friendly challenges between teams. The 2 or more teams are created based on the # of enrolled members in a program. Members are then randomly assigned to the teams using a set of Actions that process iteratively (@Steve 's miracle looping actions) stopping when the correct number of members have been assigned.
Occasionally, when I run the process, I will get a member assigned to two different teams. I am fairly certain that this is due to the processing of the Team rows happening in parallel. As I watch the rows being written to the sheet they are not in a sequential team order implying multiple processes writing rows.
Additionally, I use the same construct (see example action below) to generate the Teams themselves and I can tell these are written sequentially. So, I suspect that once the number of rows to process reaches a certain number, parallel processing kicks in.
Does anyone have any experience to confirm the above?
If this is all true, there is a HUGE assumption being made here - that each row has no dependency on the other rows being processed. In my case, I am trying to take a pool of Program members and randomly assign them 2 or more teams. If those team assignments are for the same Program and being processed in parallel, then there is a risk that both pick the same random member at the same time ending up with a member assigned to two different teams within the Program.
As I think about how to permanently mitigate this issue, I am not seeing a way around it. No matter how I re-structure the processing, if parallel processing is truly being automatically introduced based on number of rows, then there is always a chance of the number of rows becoming large enough to encounter this issue.
I need some other way to randomly assign members to teams.
Any other ideas?
Example Action that Selects rows to be processed
