So a bit of background, I'm running a little web app using just in-memory data. It's about 20 rows all-in so I've removed the database to speed things up.
Just recently I deployed it on a Raspberry Pi, mostly to see if I could. However it's thrown up an issue with NGINX deployment.
It looks like, as a remedy against memory leaks, NGINX dumps the application instance every 12 hours or so and starts up a new one.
Edit: As requested by @gf_, The result is that any in-memory data being used by my app is lost, and the re-initialisation of my environment rebuilds the default data set. The server logs show a different pid by the timestamp.
Is there a way to stop this happening? Or increase the time between occurrences? Or set a threshold for memory usage?
I'm using NGINX with the Passenger module to deploy a Rails 4 application on Raspian.