//leakcanary-repo/shark/HprofHeapGraph/Companion/INTERNAL_LRU_CACHE_SIZE
INTERNAL_LRU_CACHE_SIZE¶
[jvm]\ var INTERNAL_LRU_CACHE_SIZE: Int
This is not a public API, it’s only public so that we can evaluate the effectiveness of different cache size in tests in a different module.
LRU cache size of 3000 is a sweet spot to balance hits vs memory usage. This is based on running an instrumented test a bunch of time on a Pixel 2 XL API 28. Hit count was ~120K, miss count ~290K