I’ve tested that in fact, ongoing calls do not get killed if “core reload” gets executed in the CLI. I was wondering what the result of this is in regards to memory management: Does executing “core reload” flush Asterisk’s allocated memory while keeping current calls’s allocations? e.g. only core processes get respawned?
On the same note, is there any documentation regarding memory internals/thread management of Asterisk without resorting to source code?
Reload gets devolved to individual modules. Processes are not necessarily terminated, and some trouble is gone to to preserve data structures and their dependents, when they haven’t changed between old and new configurations.