Unpack node_exporter binary step fails on cluster

[sjpb]

What happened? Task "Unpack node_exporter binary" fails when running on a vagrant cluster in CI. I've had this happen a number of times, rerunning usually fixes it.

Interestingly I've never had the same thing happen on a real cluster.

Did you expect to see some different? Yes

How to reproduce it (as minimally and precisely as possible): Difficult!

Environment

• Role version:

galaxy latest

• Ansible version information:

4.5.0

• Variables:

• Ansible playbook execution Logs:

Looking at the source this delegates to localhost, but there's no run_once: true, so to me it looks plausible this is a race condition when running on multiple hosts?

Sorry I know this is a slightly rubbish bug report but it's the best I can do at present!

[zeroZshadow]

I've been noticing this as well. It's also running all "localhost" tasks once for every host i'm attempting to deploy too, which is quite a lot.

[0xWilliamWang]

I have the same problem and can fix it by manually decompressing the file.

[SuperQ]

This role has been deprecated in favor of a the prometheus-community/ansible collection.