I needed to run a flow that processed almost two thousand records and matched them with other client tracking records. In one branch of logic this flow could also trigger additional flows that help with data accuracy confirmation.
Usually the flow will be running in small quantities (one or two at a time) but in this instance we built it new and needed to run it on 1,500+ existing records.
Not surprisingly, this caused the flow to get Throttled.
My question is for clarification…do users need to be aware of and re-trigger anything when flows get throttled? Or will they simply be put into a queue and processed as time passes?
I’m sure I’ve seen documentation before on rate limits and when flows will get throttled, but I couldn’t find that again today when looking. What is the rate on automations running per hour?
thanks for bringing this up and for pointing out your questions in detail.
Throttling
First of all it is worth noting that throttling always happens for an entire organization, not for individual flows at the moment. This was a decision made in order to prevent any bad neighbor issues and to guarantee the behavior documented on our trust page.
The throttling will only last for a couple of minutes and will reduce flow execution throughput drastically to prevent excessive resource usage. There is nothing you need to do apart from being patient - no executions will be dropped or skipped, but they will just be queued and worked off “over time”.
Documentation
I will add a dedicated throttling page to the developer docs’ workflow automation section and post it here.
Thanks for bringing this up!
Let me know if you have any follow up questions or something needs more detailed clarification.