Deploy failing on cryptography build

Deployment: Local Node
EaaSI Version: eaasi-release-2020.03
Browser: NA
Description: After the meeting, I decided to redeploy our Eaasi node to walk through that process again. Previously, I followed the instructions here: Setup and Deployment — EaaSI documentation and things went swimmingly. Now, however, the build is failing during the with the following error:

ERROR: Failed building wheel for cryptography
ERROR: Could not build wheels for cryptography which use PEP 517 and cannot be installed directly.

I am doing this on a Centos 7 box with current updates and latest docker from the docker repos.

Are you able to reproduce the issue or did it happen once? What steps can you take to repeat the issue? What did you expect to occur and what was the actual outcome?

The steps I followed were straight out of the docs, so those should generate the same results…?

Urgency: Just a testing question and no one is gated. No rush.

Good morning @thorbrickman! Thanks for reporting this. I will flag this for @oooleg to take a look at - it may have been resolved already for the next release, but if not we’ll make sure it is.

Hello Thor! The error is caused by recent upstream changes in one of Ansible’s dependencies, which our installer is based on. That changes broke a lot of other projects too. We have fixed the installer already, but haven’t released a new version yet.

As a workaround for now, you could try to apply alpine-313.patch (1.3 KB) like this:

$ cd <eaasi-installer>/eaas/ansible
$ git apply ansible-313.patch

Then simply run again and follow the instructions from the docs.

I recently hit this issue and the patch allowed the to run without issue. I changed nothing else in my previous configuration and am now getting a:
fatal: [eaas-gateway]: FAILED! => changed=false msg: 'AnsibleUndefinedVariable: {{ eaas.public_port | default(docker_port) }}: {{ docker.port }}: ''dict object'' has no attribute ''port'''

I can start a new thread for this, if needed, but just wanted to note the side effect here. Do I need to now specific the port? It was listed as optional in the installation and a grep shows it should just be set to a docker default if not provided.

@ekaltman, which task fails with this error? Could you post the whole task’s output from the log?

@oooleg Sure, I’m running and getting the following output from that script (attached). It appears that something is amiss with the eaas-server docker initialization. I’ve also attached my eaasi.yaml (which was working previously)
eaas_csuci_deploy_out_20210505.txt (47.5 KB)
eaasi.yaml (1.1 KB)

My issue was that I apparently ran the git-pull script and was not on the right release branch, I wiped out the whole installation and re-deployed and appear to be working again.