Over the past few months I've found myself increasingly interested in Kubernetes and overall containerization of applications/microservices. I have been experimenting with many things, including some of my own apps.
However, I've always thought of adapting spigot to run in a kube cluster. Not bungeecord-esche, but rather having a single server deployed across multiple pods and acting as if it was a single running spigot server.
The storage replication would be easy to do with a DFS across multiple redundant server nodes, but the difficult part would be replicating the process identically across all pods to act as if they were a single server if players were connected to multiple different ones.
I'm still developing this at the moment, but I was wondering from any more experienced Java developers familiar with the topology of spigot whether this is technically feasible or if the effort needed to accomplish something like this would not yield tangible benefits.
However, I've always thought of adapting spigot to run in a kube cluster. Not bungeecord-esche, but rather having a single server deployed across multiple pods and acting as if it was a single running spigot server.
The storage replication would be easy to do with a DFS across multiple redundant server nodes, but the difficult part would be replicating the process identically across all pods to act as if they were a single server if players were connected to multiple different ones.
I'm still developing this at the moment, but I was wondering from any more experienced Java developers familiar with the topology of spigot whether this is technically feasible or if the effort needed to accomplish something like this would not yield tangible benefits.
