Hello,
I have just started playing with Consul and it’s amazing! Still, I’m hitting a problem:
Two VMs, each running a Consul server, each in a different DC (dc1 and dc2).
To those Consul servers’ I’ve registered two agents, with a service called nginx that’s pointing to a local nginx server running on port 8080.
The agent in Dc1 also has another service called fe which points to a different Nginx instance on port 80.
All good and well, I can see the services registered correctly, I can reach them, etc. I though, well, I should add a mesh gateway and play with that. I did it and I can reach, from Fe, the Nginx in the same cluster.
So, now I wanted to add a resolver. And this is where I’ve hit a problem. I’ve set the default service to:
Kind = “service-defaults”
Name = “nginx”MeshGateway {
Mode = “local”
}
and the resolver to:
Kind = “service-resolver”
Name = “nginx”Failover = {
“*” = {
Datacenters = [“dc2”]
}
}
If I try to reach, from FE the Nginx in the local cluster via the mesh, it works. If I simulate a problem, and stop that Nginx container, I see this in the mesh gateway logs:
Mesh gateway logs in DC1:
[2025-05-08 10:19:13.729][37446][debug][filter] [source/extensions/filters/listener/tls_inspector/tls_inspector.cc:137] tls:onServerName(), requestedServerName: nginx.default.dc2.internal.09a8af09-cb70-7ae6-edac-7704dc3cf2c9.consul
[2025-05-08 10:19:13.730][37446][debug][filter] [source/common/tcp_proxy/tcp_proxy.cc:264] [Tags: “ConnectionId”:“2”] new tcp proxy session
[2025-05-08 10:19:13.730][37446][debug][filter] [source/common/tcp_proxy/tcp_proxy.cc:441] [Tags: “ConnectionId”:“2”] Cluster not found nginx.default.dc2.internal.09a8af09-cb70-7ae6-edac-7704dc3cf2c9.consul and no on demand cluster set.
[2025-05-08 10:19:13.730][37446][debug][connection] [source/common/network/connection_impl.cc:150] [Tags: “ConnectionId”:“2”] closing data_to_write=0 type=1
[2025-05-08 10:19:13.730][37446][debug][connection] [source/common/network/connection_impl.cc:276] [Tags: “ConnectionId”:“2”] closing socket: 1
Mesh gateway logs in Dc2:
[2025-05-08 10:20:10.110][37503][debug][filter] [source/extensions/filters/listener/tls_inspector/tls_inspector.cc:137] tls:onServerName(), requestedServerName: nginx.default.dc2.internal.09a8af09-cb70-7ae6-edac-7704dc3cf2c9.consul
[2025-05-08 10:20:10.111][37503][debug][filter] [source/common/tcp_proxy/tcp_proxy.cc:264] [Tags: “ConnectionId”:“1”] new tcp proxy session
[2025-05-08 10:20:10.111][37503][debug][filter] [source/common/tcp_proxy/tcp_proxy.cc:441] [Tags: “ConnectionId”:“1”] Cluster not found nginx.default.dc2.internal.09a8af09-cb70-7ae6-edac-7704dc3cf2c9.consul and no on demand cluster set.
What’s going on ? Why isn’t that service found? What am I missing?