Sometimes I’m facing some random errors in flows, which I guess are just temporary exception.
My question is - do you have any suggestions - how to best handle restarting of failed flows?
Unfortunately the Flow doesn’t provide interaction with the failed event, so I guess there needs to be a timer embedded in the flows, which will trigger a restart flow if the flow fails to complete within a needed time.
Do you have any better ideas?
Are there any particular reasons that your flows are failing? It may be best if we look into resolving those first!
Flows can be abandoned by using keywords like ‘quit’ or ‘bye’ but this is a manual input from the user. A problem with trying to automate an exit from a flow is that the flows are run through in sequence and as such each and every action must be executed or checked - if the flow never makes it to your restart action then it won’t be able to carry it out.
Any chance you could share a bit more information about your flow?
Well, the most common error is for example “ESOCKETTIMEDOUT” when I’m using a web-hook to my service.
The whole Flow fails and I have no alert, that it failed nor can I restart it.
But there could be many mistakes I overlooked. For example I’m using Podio service and sometimes it returns me shuffled fields in output and the flow fails again.
So I need to have information that it failed. After all you post it to the Interaction stats, so why not to make an alert to a developer?
Here is an example of failed flow where I don’t understand the error:
Hi again Alexander,
Out of interest, have you seen our docs on the Podio service? Specifically the known issues section - if not, it could be worth running through the steps there.
Am I right in thinking that your Podio action in that screenshot is for update an item?
Thank you. However the point of this whole thread is to understand how to handle failed flows, not to solve particular issues.
Im inclined to try to log a start of each flow in Google spreadsheet and log the finish of the flow.
If the flow has not updated the spreadsheet with Finish log within a reasonable time, then I’ll fire an email to the Flow to run the flow again. Hm?
This is something I’m pretty sure we’ll be fixing as soon as we can - I realise it’s a problem and needs a solution. Have you see the item on our roadmap? If not, please add your vote there.
I don’t think there are any workaround for the moment.
was this flow restart implemented?