Ilya Lobkov

Another possible solution: we could avoid mounting `networkservicemesh.io.sock` using `nsmdp`. NSMgr claims where socket located in the host path. NSC/NSE mount that server socket wherever they want in container path....

It could be manually configured for any amount of clusters. But spire hasn't have appropriate API to configure Federations dynamically yet. So I created that issue to just keep Spire...

Manually means I have to add to `server-configmap.yaml` info about every other cluster: ``` experimental { bundle_endpoint_enabled = true bundle_endpoint_port = 8443 federates_with {{ .Values.federatesDomain }} { bundle_endpoint_address = "172.17.0.2"...

Method `GetEndpoint` could be called in 2 cases: 1. We are local NSMgr and we got request from NSC: in that case `targetEndpoint` is empty and we have to go...

xref https://github.com/kumahq/kuma/issues/4941

@lahabana I merged `master` yesterday, I think it didn't change since then

There are 2 cases when DPP is shown up as Offline: 1. All inbounds are offline, but kuma-dp is connected to kuma-cp (your case) 2. kuma-dp is disconnected from kuma-cp,...

Is it possible to change `controllingZone`? It feels tricky to synchronise because `controllingZone` is changed on Global CP while it's checked on Zone CP

> I think we should use a label only for 2.0 resources, no? This is what happens in PR. Only new policies will get this label. Webhook spec looks like...

@lukidzi what happens if application binds to `127.0.0.6` when acts as a client? Does it bypass the envoy in that case?