Tim Shakarian

I would focus instead on integrating with 3rd party key management solutions such as: - AWS KMS - Hashicorp Vault - Google KMS This will solve the disaster recovery scenario...

The challenge here is that people who don't have access to any form of secure credential storage outside of k8s are not likely to be equipped to securely store a...

For me I'd love a similar solution to `minikube tunnel`. I test multiple services exposed via an istio's ingress-gateway and use DNS for resolution with fixed ports. The DNS config...

Are you using a shallow clone? If so shallow clones are broken (https://github.com/src-d/go-git/issues/1143). Even without shallow cloning, this can still happen on occasion. This appears related to https://github.com/src-d/go-git/issues/1151. Here's my...

I'm having the same issue as well. I'm in Slack and happy to help debug or problem solve.

From a safety perspective I actually like this behavior. Maybe the fix is to just improve the error message and (not mutually exclusive) implement the attribute based deletion.

We certainly want to avoid deleting resources in this case. However, I strongly prefer a partial apply as opposed to aborting. The behavior should be to skip the problematic file(s)...

To accommodate this I would recommend adding an optional field like `defaultNamespace` to the `GitTrack` spec. If `defaultNamespace` is not specified then the resource should not apply. That said, the...

Thanks for the detailed response. I think you made a strong case for `--namespace` being sufficient.

After some thinking there is one downside to the `resource-state` annotation approach: If someone were to `kubectl apply -f` the git folder it would recreate deleted items. An alternative may...