I’m currently working on a tvOS app. This is my first native (Swift) app. The app will be a digital signage app, used during events or in offices of companies. One big difference compared to a typical app on iOS/tvOS is that it needs to run pretty much 24/7, so memory is a big topic for this app. The smallest leak will eventually cause the app to crash.
The app is constantly looping through a set of fullscreen slides. At the bottom of the screen there is a ticker with 10 articles (refreshed every 10 seconds - now during development). Below is a screenshot of the weather slide, to get an idea.

Currently the app is crashing after a period of time and I’m pretty sure I’ve narrowed it down to the ticker component (when disabling it, the app lives for days). If I use the ‘Leaks’ preset in Instruments I get the following result:  It looks like it’s leaking Article instances. I’m recreating Article instances every 10 seconds and providing them to the ticker component. I think that is why new instances leak every ~10 seconds.
Before I started using the ‘Leaks’ preset in Instruments, I used the ‘Allocations’ preset, while using this all seemed fine to me. But I’m probably misreading the results…
Using allocations:  The way I read this is that currently 10 Article instances exist in memory, and 31 have existed but are cleaned up now - so I’m safe.
But the app still crashes.
I’ve read a lot on retain cycles, implemented weak/unowned where I believe I should.
So my question is not so much about code, but more about how to read this data, what does a Leak mean in this context, and why do I see these ‘leaks’ not as persistent objects in the Allocations window?
(tests are done on multiple devices + simulator)