Update Gemini config for https://github.com/Islandora/Crayfish/pull/109

[ruebot]

GitHub Issue: Islandora/Crayfish#109

What does this Pull Request do?

Update Gemini config for Islandora/Crayfish#109

• Remove crayfish_gemini_fedora_base_url
• Add crayfish_gemini_fedora_domain
• Add crayfish_gemini_drupal_domain

Hit this problem this afternoon building a new machine:

[2020-12-02 18:29:58] app.CRITICAL: Pimple\Exception\UnknownIdentifierException: Identifier "crayfish.drupal_domain" is not defined. (uncaught exception) at /var/www/html/Crayfish/Gemini/vendor/pimple/pimple/src/Pimple/Container.php line 101 {"exception":"[object] (Pimple\\Exception\\UnknownIdentifierException(code: 0): Identifier \"crayfish.drupal_domain\" is not defined. at /var/www/html/Crayfish/Gemini/vendor/pimple/pimple/src/Pimple/Container.php:101)"} []
[2020-12-02 18:29:58] app.DEBUG: < 500 [] []

Interested parties

@seth-shaw-unlv @dannylamb

[seth-shaw-unlv]

Smoke tested with no issues. 👍

[seth-shaw-unlv]

Travis had an issue downloading a part. I've restarted the build. Hopefully it will work this time.

[seth-shaw-unlv]

@dannylamb, Travis is failing here because ansible-galaxy is still attempting to install the "master" branch (which no longer exists). I don't see a way to tell galaxy to change this behavior. The tests do have their own requirements file, but it isn't structured like the playbook where we can specify a version/branch. 😕

Any ideas on fixing this?

[dannylamb]

@ruebot Do you think you could put this chunk in tests/requirements.yml for the keymaster role. I think you just need to explicitly declare that you're using the main branch and it'll all sort out.

[ruebot]

@dannylamb sure thing!

[seth-shaw-unlv]

Travis is taking it's time getting around to testing this. I'll keep my eye on it.

[seth-shaw-unlv]

Travis failed again. Apparently we don't have git on the test image?

[WARNING]: - Islandora-Devops.keymaster was NOT installed successfully: could not find/use git, it is required to continue with installing https://github.com/Islandora-Devops/ansible-role-keymaster

[dannylamb]

🤦‍♂️

[ruebot]

@dannylamb @seth-shaw-unlv I was looking at the TravisCI config files to figure out the testing infrastructure on the Islandora-Devops roles. Looks like a lot, if not all of them are on 16.04. Do they need to need to be updated to at least 18.04 to match with the default configuration in the vagrant file? Though, I'm not sure if that would address the more immediate concern of git being installed.

[seth-shaw-unlv]

@ruebot, worth a shot!

[ruebot]

I just realized I was pushing to the canonical repo, instead of our fork. Sorry about that y'all. Really embarrassing on my part, and glad I didn't push to main 🤦‍♂️ 🙏

[dannylamb]

main is protected, but anybody can push a topic branch to our canonical repo for testing. No big deal!

[seth-shaw-unlv]

@dannylamb & @ruebot, looking at the failure some more... @ruebot updated the tests to use 7.4, but it appears the Crayfish composer lockfile is imposing constraints that prevent it from running on 7.4 due to dependency issues, especially with the move to Composer 2.

So, we need to get a Crayfish PR that brings things up to 7.4... which now makes it part of the D9 push.

[ruebot]

@seth-shaw-unlv that's my end goal for Islandora/Crayfish#110 now. Just waiting on the Chullo PR to sort out, and then I'll jump on this one.

[seth-shaw-unlv]

As long as Travis comes back green, this will be good.

[ruebot]

I think we're getting the Idempotence test: fail because of these two spots on the second run:

1. https://travis-ci.org/github/Islandora-Devops/ansible-role-crayfish/jobs/749705936#L1374

   • this has a force, so if I understand ansible correctly, that will always fail an Idempotence test because it is a force.

2. https://travis-ci.org/github/Islandora-Devops/ansible-role-crayfish/jobs/749705936#L1438-L1441

   • Similar to the above, I think the way the "Template Symfony config files" task is written will also always fail an Idempotence test because anytime one of the items in the loop is a file, it's going to modify it. So, it's going to always be changed.

@seth-shaw-unlv @dannylamb lt me know how you want to handle this. I'm not really around until January 7th. But, happy to push this across the finish line if y'all want further action.

[seth-shaw-unlv]

I thought you were already gone! I was just looking at this. I think we will want to get these two bits resolved, but since we can push to your branch we should be able to work it though. Go enjoy your vacation! 😀

[seth-shaw-unlv]

Huh, we've been ignoring idempotency since 2782ab0 back in February when the force flag was added for 're-runability'. So, I guess if everyone else is fine with it, we will run with it. We can open a new issue if folks decide we need to revisit this.

You've done well, @ruebot, sir.

[ruebot]

Thanks @seth-shaw-unlv 🙏