[COST-4981] reset local unleash

[maskarb]

Jira Ticket

COST-4981

Description

Closes #5068

This change will:

• hard reset unleash locally:
  • remove everything in the flags.json (this is all deprecated anyway, so we should not rely on this method to seed the db)
  • create 2 "template" flags: cost-management.backend.schema-flag-template and cost-management.backend.source-uuid-flag-template
  • these flags can be cloned to create a real flag for local testing
• move the .unleash dir to dev/containers/unleash
  • update the docker-compose to use this new directory
• with unleash v5, we need to use a personal access token (PAT) to use the API. Unfortunately, you need to have a PAT in the db in order to use the API to create new PAT. To get around this, I've written a python script that directly seeds the unleash db with a PAT (dev/scripts/setup_unleash.py). This script also seeds the template flags.
• update unleash version to match the insights instance
• several new env variables:
  • the old admin unleash token is renamed UNLEASH_ADMIN_TOKEN
  • the new UNLEASH_TOKEN is scoped to the development environment
  • UNLEASH_PAT is the user PAT used to seed the db and can be used with any API calls wanted

Testing

1. make docker-down delete db
2. make docker-up-db
3. see logs:

$ make docker-up-db
/Users/mskarbek/.rd/bin/docker compose up -d db
...
[+] Running 1/0
 ✔ Container koku-db  Running                                                                                                                                                                                                                                       0.0s 
 PostgreSQL is available!
/Users/mskarbek/.rd/bin/docker compose up -d unleash
...
[+] Running 2/0
 ✔ Container koku-db       Running                                                                                                                                                                                                                                  0.0s 
 ✔ Container koku-unleash  Running                                                                                                                                                                                                                                  0.0s 
dev/scripts/setup_unleash.py <<<<<< all of this is new
waiting for unleash to become ready
unleash is ready
adding unleash personal access token to db
[('user:6188b62f2f59348f3c195b66983147111682f4bb78a3f7ed9626bd84', 'admin', 1, datetime.datetime(3000, 1, 1, 0, 0, tzinfo=datetime.timezone.utc), datetime.datetime(2024, 6, 5, 15, 16, 14, 82000, tzinfo=datetime.timezone.utc), datetime.datetime(2024, 6, 5, 15, 13, 17, 56355, tzinfo=datetime.timezone.utc), 1)]
creating template unleash flags
200
...

4. log into the unleash instance and see that there exists a PAT:
   http://localhost:4242/profile/personal-api-tokens
5. whilst in unleash, see that there are only 2 unleash flags (mentioned above)

Release Notes

• proposed release note

* [COST-####](https://issues.redhat.com/browse/COST-####) refresh unleash locally

[samdoran]

It's possible this could be replaced by a health check in the compose file. The unleash service could benefit from adding a healthcheck for the db service as well.

[maskarb]

What would the healthcheck be for unleash?

[samdoran]

The unleash image only has busybox, so something like this.

healthcheck:
  test:
    - "CMD"
    - "wget"
    - "-qO-"
    - "http://localhost:4242/health"

That still needs some refinement.

[maskarb]

I don't know how to leverage a healthcheck here. Even when adding one, the make command just moves on, healthy or not. The DB healthcheck is good though, so I added that. I'm just not sure we can do away with the wait_unleash_ready from this script.