I am using Swarm 3.0.5 with libcluster. Imagine, I have two nodes starting up.
On every node startup same procedure is invoked:
Swarm.whereis_name to see if name is registered
if not, start GenServer using via tuple
try to register with Swarm.register_name/2
NodeA starts up and follows procedure and registers new process
NodeB follows same procedure, but as soon as Swarm application starts on the NodeB, before even proceeding to my custom logic, on NodeA I get the following in debug log level:
topology change (nodeup....)
removing registration for XXXX, nodeA is down
even though nodeA is up and after that I can run Swarm.registered/0 and see process registered on NodeB.
Hello!
I am using Swarm 3.0.5 with libcluster. Imagine, I have two nodes starting up.
On every node startup same procedure is invoked:
NodeA starts up and follows procedure and registers new process NodeB follows same procedure, but as soon as Swarm application starts on the NodeB, before even proceeding to my custom logic, on NodeA I get the following in debug log level:
even though nodeA is up and after that I can run Swarm.registered/0 and see process registered on NodeB.
Any thoughts about this behavior?
Thanks!