Andrew Gerrand

> Having a docker image would allow me to easily run upspin on Google Container Engine (GKE), a custom-hosted kubernetes instance or a more bare-bones setup like a CoreOS based...

@awly IIUC LetsEncrypt requires that you be listening on port 443 to issue certs.

@swetland once you have the Docker container how do you run it? I agree systemd and xinetd are icky, but what's the replacement for those things?

> You’d run it the same way you run any other container: using fleet, kubernets, swarm, systemd, $favorite_tool_here. Why are you asking? :) So how is that any less onerous...

@awly yes, please! I used `xinetd` because that was what I knew. If there's a simpler way, let's do it. Would you mind sending the doc change?

We should definitely do this. It would let us automate the creation of the Google Cloud Compute (or AWS EC2) instance that runs upspinserver, and therefore automate most of the...

I've made an image available publicly as `gcr.io/upspin-containers/upspinserver`, and is prepared by [upspin-deploy-gcp](https://godoc.org/gcp.upspin.io/cmd/upspin-deploy-gcp/). I'm now working on automatically updating that image.

The image is now automatically updated on each push to the gcp repo. All that's left to close off this issue is to document the existence of the docker image.

Did it stop doing this after you restarted it? Or is it reproducible? You might want to run the server with the environment variable `GOTRACEBACK=2` set, and then you can...

I think the root issue is that the `upspin.io/dir/server/tree.Tree` type has a String method that does a nontrivial amount of work to stringify the Tree. This involves taking a lock,...