{ TimeoutError: Event response for "sccBrokerJoinCluster" timed out
at Timeout.setTimeout [as _onTimeout] (/usr/src/node_modules/socketcluster-server/serversocket.js:1238:19)
at ontimeout (timers.js:436:11)
at tryOnTimeout (timers.js:300:5)
at listOnTimeout (timers.js:263:5)
at Timer.processTimers (timers.js:223:10)
name: 'TimeoutError',
message: 'Event response for "sccBrokerJoinCluster" timed out' }
How many nodes in your cluster approximately?
What sort of CPU usage do you see on each instance?
Where did you obtain this log from? scc-state or scc-broker or from your socketcluster instance (scc-worker)?
I would check for version compatibility; in SCC, the major version number of all the components must match
So for example, scc-broker v8.x.x must be paired with scc-state v8.x.x
Maybe you need to add more scc-broker instances to your cluster. Could be that they're so busy that they're timing out and you need to add capacity
I would also check scc-state CPU usage
It could be any of these things
State Server: socketcluster/scc-state:v8.0.0
Broker: socketcluster/scc-broker:v8.0.0
Environment:
SCC_STATE_SERVER_HOST: scc-state-service
SOCKETCLUSTER_WORKERS: 1
SOCKETCLUSTER_BROKERS: 1
SCC_INSTANCE_IP: (v1:status.podIP)
SCC_BROKER_SERVER_LOG_LEVEL: 2
Ratio: 1 state server to 4 brokers
All running on AWS EKS
The log snippet came from the state server
Checked connectivity from broker -> state via
telnet scc-state-service 7777 and it connected just fine
on all instances
of the broker server that is ^^
SOCKETCLUSTER_WORKERS: 1
SOCKETCLUSTER_BROKERS: 1
SOCKETCLUSTER_BROKERS: 1
should those be set or left unset for any environment !dev?
it looks like the broker is unable to reach the scc-state server, since it's waiting for a response back but it times out, it might show that the connectivity works via telnet but it could be configuration issue with how the broker is trying to resolve the host
Hi there! I'm wondering how one can produce/trigger an event/rpc from another event/rpc :question:
The reason I want to do that is to allow several sources to trigger same event or proxy events from external sources to internal events.
E.g. some logout event can be reported by external system via Redis Pub/Sub or REST API, but I want that trigger existing event handler for this operation. It may seems not straightforward or maybe bad example but I have a case where I am subscribed to redis pub/sub for certain event and I want that event to trigger a native SC event.
Or is this done through some other method besides sockets?
put into the phone app all 3 components and configure leader election for scc-state
but it is a bigger project i think there is nothing already coded like leader election or hot standby for scc around
that is also possible as you have control over the network you could get a fixed list of client-ip's
no one did it befor