Hi everyone,
I’m setting up a new docker compose stack to run concentratord, mqtt-forwarder, chirpstack and a few more applications.
To get mqtt-forwarder and concentratord to communicate, and to have each running in their own docker container, I modified the configuration to use TCP instead of IPC for zeroMQ (which is one of the great things of zmq).
[concentratord.api]
# Event PUB socket bind.
event_bind="tcp://*:3000"
# Command REP socket bind.
command_bind="tcp://*:3001"
However, I don’t think it’s something concentratord is particularly happy with, and it’s crashing as it’s unable to rebind the socket when reloading because of a configuration change.
2025-02-19T11:16:15.197Z INFO [libconcentratord::jitqueue] Initializing JIT queue, capacity: 32
2025-02-19T11:16:15.197Z INFO [libconcentratord::events] Creating socket for publishing events, bind: tcp://*:3000
thread 'main' panicked at chirpstack-concentratord-sx1302/src/cmd/root.rs:66:63:
bind event socket error: Address in use
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
INFO: Configuring TTY
INFO: Flushing TTY
For the time being i’ve made a named volume that i can use to share the ipc sockets between the two containers but it would be nice to simply use zmq TCP.
Just wondering if someone got it working over TCP and it’s something I’m doing wrong, or if it’s just not supported at the moment by concentratord?