Jeff Lill

I just noticed wile debugging in the **neon-node-agent** pod container that there are a bunch of environment variables that shouldn't be there, like: ``` NEON_DASHBOARD_PORT=tcp://10.253.133.192:80 NEON_DASHBOARD_PORT_80_TCP=tcp://10.253.133.192:80 NEON_DASHBOARD_PORT_80_TCP_ADDR=10.253.133.192 NEON_DASHBOARD_PORT_80_TCP_PORT=80 NEON_DASHBOARD_PORT_80_TCP_PROTO=tcp NEON_DASHBOARD_PORT_9762_TCP=tcp://10.253.133.192:9762...

build: don't publish tools like neon-cli when nothing has changed

2

Several of our projects include a **PubCore** build target that executes after the build. This is quite slow and unfortunately happens even when the project didn't need to be rebuilt....

I think this will need to take a special command line option that identifies target cluster's definition since there won't be a current cluster in this case.

Fix hardcoded Azure regions

1

The Azure `SubscriptionResource.GetAvailableLocationsAsync()` method does not appear to be working; it always returns an empty collection. This may be permissions problem, but I would have expected to see a security...

- [ ] Identify the required Azure permissions and write a PowerShell script to create an app service principal we can use to provision clusters. https://docs.microsoft.com/en-us/azure/active-directory/develop/howto-create-service-principal-portal

Now that basic single-node cloud cluster deployment works, we need to do some additional testing. This is necessary because clouds (Azure specifically) has some non-obvious constraints around availability sets, proximity...

We'll need to support node scale-out etc. for cloud environments eventually as well as other integrations such as [LoadBalancer](https://kubernetes.io/docs/concepts/services-networking/service/#loadbalancer) services. We haven't really drilled into this but I suspect that...

Optimize cluster GetClusterStatusAsync() methods for cloud?

1

We're calling ` IHostingManager.GetClusterStatusAsync()` every 5 seconds right now from neonDESKTOP to update the task bar icon state. This is quite costly for **AWS** because it requires listing all cluster...

We're currently having **podman** use the **root** user credentials to authenticate with the local Harbor registry. This is broken because the root user will likely change their password sometime in...

Preview release requirements: - [ ] prompt for cluster sso password - [ ] create AWS app credentials - [ ] identify minimal required permissions - [ ] neon-cli -...