How are you queuing your API Workflows?

I’ve been searching through the documentation and I can’t seem to find whether API workflows are queued. I’ve read some forum posts saying they are, but can’t confirm to what extent.

For those of you who work heavily with api workflows, how are you managing them? For instance, what methods are you using to queue requests. Zapier? Amazon Simple Queue Service (Amazon SQS)? Database?

Really curious about how people are handling this, considering the API workflow system can be unreliable at times.

To the extent it’s helpful, I’m pretty sure API workflows are run in the background and are deprioritized relative to user-facing actions (e.g., page loads, conditions, workflows/actions). However, this only matters if your server hits 100% capacity. If it never hits 100%, then prioritization decisions don’t need to be made since the server can handle it all at once.

1 Like

Hmmm that’s good to know. I guess right now I just have to focus on working around errors during the API workflow such as this: Missing things omitted from list race condition

Server logs at least provide a good overview of which workflows ran and which didn’t, so will work from there. Thanks!

Well… I am using some workflows… and thought about using Zapier. But since I only have few events where I need a scheduled workflow I am managing them via user triggered events of users visiting my site. In order not to generate a lot of unnecessary workload, I have the execution of the workload triggered to a time based condition. Means: My workflow should only be executed 1x every 12 hours… so the workflow includes a step of maintaining a field within my database with the time it was last executed… and the workflow initiation is based on the condition that this time lies more than 12 hours back :slight_smile:

1 Like