Thanks for sharing this feedback and the specific scenario — that context is really helpful. I want to make sure I fully understand how you would like this to work.

In your ideal world:

• Would compact run on a set schedule as part of your regular maintenance window, or only when certain conditions are met (for example, when fragmentation or oplog size crosses a threshold)?

• Should compact run automatically every time you have a maintenance window, or would you prefer it to be an optional step you can enable/disable per window?

• Since compact can consume time and resources, how would you want us to handle the performance impact — for example, by setting time limits, resource limits?

• Instead of automating, what if we provided a self-service control for this (for example, a button in the UI to trigger an ad hoc compact, or the ability to script it via an API) versus us handling it entirely as part of an automated maintenance policy?

Any additional details you can share on your preferred behavior or safeguards (timeouts, limits, notifications, etc.) will help us evaluate the feedback here.

In the meantime, you should always be able to trigger the compact() command on a database anytime yourself.