Production incident response for crashed services almost always starts with `gdb ./binary core` and a `bt full`, since a coredump captured on a crashing server preserves the exact memory state a live debugger could never catch after the fact. systemd-coredump and Linux distributions configure core_pattern precisely so on-call engineers can do this post-mortem triage instead of guessing from log lines.