(experimental feature)
Before a worker can start, it should have a configuration file. In case if such a file is not found it may request configuration from the coordinator to which it connected.
Added example logic if a worker is needed to be blocked until a successful packet exchange with a coordinator is being made.
* Add error return for config loader
* Add config loaded flag to worker
* Add zone flag
* Add a custom mutex lock with timout
* Refactor worker runtime
* Refactor internal api
* Extract monitoring server config
* Extract worker HTTP(S) server
* Add generic sub-server interface
* Add internal coordinator API
* Add internal routes and handlers to worker
* Add internal worker API
* Refactor worker run
* Migrate serverId call to new API
* Add packet handler to cws
* Extract handlers for internal worker routes in coordinator
* Pass worker to the worker internal heandlers
* Cleanup worker handlers in coordinator
* Add closeRoom packet handler to the API
* Add GetRoom packet handler to the API
* Add RegisterRoom packet handler to the API
* Add IceCandidate packet handler to the API (internal and browser)
* Add Heartbeat packet handler to the API (internal and browser)
* Rename worker routes init function
* Extract worker/coordinator internal ws handlers
* Update timed locker
* Allow sequential timed locks
* Add config request from workers
* Add nil check for the route registration functions
- All voice is fan-in and broadcasted.
- Currently, voice is sent in the same game track. TODO: send voice in different track (Having issue with dynamically addTrack after offer and answer)
* Refactor input poll to make it x2-3 faster (0.4ms+ -> 0.1ms)
* If Server is sendrecv transceiver, client needs to be sendrecv on Firefox.
c641b530a0
Co-authored-by: giongto35 <giongto35@yahoo.com>
