Best practise on automation when creating issue's

[Jeroen Gouma]

Currently thinking about how to handle the automation when creating new issue's in JSM.

We receive somewhere between 500 and 750 new issues per day, mostly via email and about 10-20% via rest API.

On the majority of the newly created issue's some automation needs to be done (filling fields based on reporter, assigning to specific worklflow, linking,  etc.).

Currently I see 2 approches:

1. Create 1 big automation rule which executes on every new issue and has several branches which are executed based on specifc values.
2. Createmore then 20 small automation rules (one for each variant) and have them all executed on creation of a new issue.

What is the best practise (and why)? 1 big automation rule or more then 20 rules which are all executed every time an issue is created? Or maybe a combination?

Please share your thoughts about this.

Kind regards,

Jeroen

[Nic Brough -Adaptavist- · Accepted solution]

I would lean towards a single monolithic automation, sort of.

The reason for this is that automations may not run exactly when you think they will, and you may find some of your automation has dependencies on earlier ones (by the way, you "set the workflow" by selecting the issue type, not with an automation)

I would also recommend doing as much as possible in post-functions on the "create issue" transition in the workflow, which are simpler than automations, and will run before the automation (so if you can set a field, or assign it in there, your automation has less to do)

[John Funk]

Hi Jeroen,

I agree with Nic - I prefer a single rule in most cases.