When working with multiple enabled workflows, each with different triggers, steps, loops, and branches, and each executing multiple times every day, you are bound to end up with a large number of workflow execution objects.
The challenge
In situations where you need to triage an issue or trace a workflow execution for a specific scenario that occurred in the past, the filtering options available in the UI today are quite limited. At the moment, filtering is primarily based on date and time, with end workflow result as an additional filter.
This becomes even more challenging when you want to determine whether a specific step in a workflow was executed for:
a specific user
a specific access request
This is especially difficult in workflows with nested branches, where finding this information through the UI often means manually sifting through multiple execution records.
Possible workaround
There are REST APIs that can be leveraged to retrieve workflow executions and apply more targeted filtering. However, that means building and maintaining a custom solution outside of the default offering.
Looking for ideas
I am curious how others are handling similar requirements.
If there are easier approaches than going down a fully custom path, I would love to hear about them.
I use the send email action to find certain scenarios or notify me of certain edge cases occurring.
My flow is typically to use a compare strings/numbers/boolean to look for a certain edge cases or criteria then I will use a send email action with the workflow execution id in the body of the email with a description of what exception occurred.
I’ve found it much easier to search my email by a subject line and copy/paste the execution id than manually digging through the logs.
Let me know if you have any questions/concerns about this approach!
I love the idea. This is a good way to be notified of errors, in low-volume use cases. I however am vary of this approach causing inbox flooding in case something goes really wrong and too often.
Understandable! I use a forwarding rule to it’s own specific folder based on the subject name to avoid it clogging up my main inbox. It’s worked pretty well for me, but maybe someone has a better way
I agree, and for anything critical, it makes a lot of sense to be notified immediately.
But I’m just wondering if the trade-off here is that even if something not super critical fails, and fails too often, we’re potentially going to run into a ton of emails. You’re then left with a similar problem: What email to search for the error (it sure is easier than looking into execution JSONs which are not searchable at the moment, but a minor inconvenience).
Even if there is a dedicated email forwarding rule for a specific workflow, or let’s say you’ve gone granular to keep specific step errors of specific workflows in sub folders, there’s still a ton of email notifications in a worst-case scenario.
Overall, I like @trettkowski’s idea because it reduces going full custom, but I really wish there was something OOTB for this…
