We have a workflow that leverages MS Power automate to send emails for every leak recorded by field staff in GIS. This fires once per hour, sending 1 email for each new leak recorded in the past hour. Last night a single new leak was entered into GIS at 10:26pm. MS Power automate shows a single email, containing this payload, was sent at 11pm as expected. However, in our inboxes we got 2 duplicate emails approx 5 min apart.
Is there a way to determine why this happened? How can I figure out if this is a hiccup at either the MS Power Automate end or our own MS Exchange end??
Hi @ahargreaves_FW - I am not sure from the Microsoft Exchange side, but I do have some tips on how I debug my cloud flows.
The first place I check is in 'Monitor' and look at the 'Cloud flow activity' this shows me all of the activity for my Power Automate environment.
If that does not give me any meaningful information, the next place I check is the Power Platform Admin Center. It provides fine grained details on the errors and activity for the individual flows.
I hope it helps.
Thanks @SeanKMcGinnis . We've had more issues with flows failing to trigger. I can easily document the creation of new AGOL features but no associated flow triggering to alert users. The various MS Power Automate dashboards and flow run history show no errors, but also no activity when those features we created. It's as if there was a Power Automate outage. But i have no way to prove this.
@ahargreaves_FW - It is difficult when working across two service-based platforms to trap where the error is occurring.
Have you set up any other testing tools to try and catch the webhooks when triggered? I use Webhooks.site to add a listener for webhooks to help debug when they fire and the corresponding payload. If there was a Power Automate outage, they post messages in their Service Health Status page. I have a Power Automate trigger listening for changes in the RSS feed to push me notices when there is a service interruption.
I apologize cause the points above don't provide a direct answer to your issue or the forensic info to look backwards, but I hope it provides some additional tools to help debug and track where the issue may be coming from.