I am seeing this pattern in RAM usage on my HA core running as a docker container on Intel NUC i5 8GB. Is there a way I can debug this?
Restart Home Assistant in Safe mode.
If the the memory is stable in this mode then a third party integration is at fault. You can work out which one by enabling them using a binary search method (half at a time so you know which half has the leak, halve again, repeat).
If the memory leak persists in safe mode then it is either an issue with a core integration or more likely you have an infinite loop somewhere in your automations or scripts.