Options to monitor Fusion scenarios that are unexpectedly down

[cverges-medallia]

Over the holidays, we had a Fusion scenario that was stopped due to a transient but long-running error in a third party service.  This caused the scenario to stop.  Unfortunately, in the slew of other scenario errors that were triggered, we missed reenabling that one scenario.

While this was our fault, we are looking to Fusion for ways to help minimize the impact of such situations.  Critical scenarios would ideally have a setting to NOT be auto-disabled, ever.  Or we could have a way to have one scenario check on the status of others and send a reminder email/Slack/etc. to reenable if it's detected.

Other than sheer will, does anyone have any suggestions of how to make this generally more robust?

[lgaertner]

Hello,

without having set up something like that in detail, my first idea would be to add a router and use a fallback route. There you could setup a notification in case of the scenario running into an error.
On the other side you need to check, if this will work out for the error type which was responsible for stopping your scenario...

Regards

Lars

[cverges-medallia]

Thanks for the reply.  I'm aware of how to add error handling for a specific module with routers for various paths of handling.  What I'm trying to avoid is needing error handling on EVERY module.  These can be very complicated and large scenarios and adding error handling everywhere makes the modules cumbersome to maintain.  As such, I'm trying to find a way to simply not allow the scenario to disable itself or to externally alert if a module is disabled or stays disabled for an extended period of time.

This may end up begin a thread about a product feature request, but I was hoping maybe someone had a thought on how to do this using current tech. For example, does Fusion itself have an API that allows one scenario to introspect status of another scenario?  Can the "consecutive errors before deactivation" be set to something akin to infinity?

[KatherineLa · Accepted solution]

Have you explored the options in Scenario Settings yet? There's some overall 'how should this scenario behave if there's an issue' settings that might help with your challenge. I think 'Allow storing incomplete executions' may help you to start, along with considering how many failed executions are permitted before a scenario completely stops itself. If you're handling significant amounts of data, consider 'allow data loss' as well but that's also heavily dependent on what kind of data you're processing and if that loss is acceptable to protect a scenario overall.

https://experienceleague.adobe.com/docs/workfront/using/adobe-workfront-fusion/scenarios-in-fusion/s...

[cverges-medallia]

Thanks, yes, I'm aware of this.  We've updated the "disable after X errors" value from 3 to 9999.  Hopefully this will help reduce the impact of events.

There is still a functionality/feature gap around setting a scenario to NEVER be disabled and/or monitoring the Fusion scenario's state via APIs or other means.  Maybe if a Product Manager is watching this thread.... (fingers crossed!)

Thank you!