Hi @Kenan qureshi -- Welcome to the Atlassian Community!
With a more complicated / large scale / heavy usage of automation environment, it is likely managing the effective use of rules (and their health) will require staffing for a full time role / position in addition to any existing Site Admin positions.
Actively manage site permissions. Work with your Jira Site Admins to audit and effectively manage who has elevated site permissions and who can add / remove marketplace addons for the site; both can greatly impact the behavior and health of automation. Same thing for JPD admin / creator or JSM admin / agent users as those have independent automation rules with special privileges that could impact the Jira software's automation and projects.
Use and enforce proper change management for rules: consider your rules to be "enterprise production code", so proceed accordingly. This will include extra work to design, test, and document rules. (Sadly, there is no REST API yet for rule management to help.) There are several community posts describing methods to use Confluence for such rule management, including bulk parsing of exported rules to create documentation pages. Change management may include using a formal intake process to define rules beyond single project scope for impact assessment, and severely constraining who is permitted to use the Send Web Request action (as that can bypass rule scoping). Perhaps also use sandboxes or free license test sites to manage the "blast radius" of rule testing and improve experimentation safety.
As @Dick describes, use the narrowest possible scope needed for each rule. Do not make them all global or multiple-project scope initially. That will certainly lead to a possibly unrecoverable / irreversible disaster due to a simple accident such as a typo. Perhaps even consider implementing automated site backups using custom code or marketplace apps to mitigate the above risks.
Read the documentation, and learn how racetrack errors / conditions manifest in rules, and learn how to mitigate them through changes in rule design and effective use of the Re-fetch Issue and Delay() actions.
Actively monitor rule usage and errors at a site level. As noted earlier, this will take time and so should be staffed accordingly.
Actively monitor the automation ecosystem by: checking the Atlassian status pages, watching the community for early-warning of symptoms / changes, reading the weekly release blog, and reading the public backlog for new defects / suggestions.
Kind regards,
Bill
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.