What's your process for debugging high-memory Node services in production?
Heap snapshots are heavy and can crash the process if taken wrong. I've used --inspect with a pause, but that means downtime. How do you profile memory without making the problem worse?
29
0 comments