instructions for standalone etcd are failing with "oci runtime error..."

[Daniel Petersen]

Hi Folks,

I'm attempting to get a standalone etcd up and running on coreos so I can further eval kubernetes. I'm following the instructions found here:

https://coreos.com/etcd/docs/latest/v2/docker_guide.html

and pulling the latest coreos/etcd image (which is currently v3.1.5)

but the container fails to run with the following error:

docker: Error response from daemon: invalid header field value "oci runtime error: container_linux.go:247: starting container process caused \"exec: \\\"-name\\\": executable file not found in $PATH\"\n".

I did have success running an older version (v2.3.7) but this version is unsupported and I'd really rather be running the latest and greatest as it's likely to play better with Kubernetes.

Does anyone know how to resolve this? I would assume that the doc just hasn't kept up with changes in the image and there's a paramater or something that needs to be updated, but I don't know where to start; any help appreciated!

Also, is there a coreos feedback channel where I should bring up issues like this so documentation can get updated?

//Daniel

[Rob Szumski]

It looks like your docker run command is detecting the etcd configuration flags (starting with -name) as the path to the binary within the container. I’d guess that the 2.x dockerfile specified this path, and the 3.x one doesn’t, _or_ your command has extra whitespace or some other malformed argument.

The etcd docs are contained within the repo, so you can file bugs or even better, contribute back fixes: https://github.com/coreos/etcd/issues

-  Rob

--
You received this message because you are subscribed to the Google Groups "CoreOS User" group.
To unsubscribe from this group and stop receiving emails from it, send an email to coreos-user...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.