Why is it taking so long to connect to my Minecraft server via MetalLB?

I've managed to run multiple Minecraft servers on my K8s cluster with MetalLB without any issues. My setup uses distinct loadBalancer IP pools assigned from a separate network interface. Perhaps checking your switch settings for gARP issues could provide clarity? Flushing the ARP cache might help too.

It would be helpful to know more details about your Kubernetes cluster. Using BGP with MetalLB might not solve your issue since Minecraft typically runs on a single pod. Have you tested MetalLB with a simple service like nginx to see if it behaves differently? That could clarify the situation.

Before investigating further, I suggest a solid diagnosis. First, test your connection bandwidth with iperf between your devices. Two minutes might indicate a DNS issue, so try using IP addresses directly instead of hostnames. Check logs as well; they can be quite revealing. If you're expecting a one-size-fits-all answer, remember that network issues often require detailed digging!

It sounds like your issue might be related to your network setup. When I had a similar experience, it was due to network topologies. If you're connecting through a router with a complicated setup or across different subnets, it can slow things down significantly. Can you share more about your network configuration? That would help us narrow down the problem.

By the way, how are you deploying your Minecraft server? Is it setup to scale? Many engineers find scaling a Minecraft server quite challenging! It’s possible to shard the server across multiple instances, but the details can get pretty niche.