Add optimize to helm charts

[Zelldon]

This PR contains:

• Add a new sub-chart for optimize (most resources were copied from Tasklist/Operate)
• Add default values to the values.yaml file
• Integrate Optimize with Identity (add configuration on both sides)
• Disable optimize if identity authentication is disabled
• Add template tests

Manual testing:

After installing we see:

Doing the necessary port-forwards we can login

Next Steps:

• We have to solve the issue with the partition count, which is necessary for Optimize Find a good way to configure partitionCount in Optimize #286
  • Per default optimize will now only import for one partition
• We will add the values documented to the README Add optimize to README  #284
• We will enhance our e2e/integration test - login to optimize Add Integration test for Optimize #283

Reviewers:

I know it looks like an big PR, but most of the lines are due to tests and golden files, which you can skip. :)

@oleschoenburg @npepinpe whoever has more time please have a look at the PR
@RomanJRW could you please check the configuration to whether this makes sense in your opinion?

related to #126

[Zelldon]

@RomanJRW it seems that Optimize only supports CLIENTID the other tools, like tasklist and operate use (or support) CLIENT_ID Is this expected? I stumbled over it, because it didn't worked without the right settings here.

[RomanJRW]

Thanks for the heads up. Once JIRA wakes up, I'll create a ticket for us to change this in future 👍

[npepinpe]

🚀

• 🤡 Kind of weird that we can't use a single .helmignore for all the charts together 🤷‍♂️
• At this point, I feel like the secret helpers might be, uh, abstracted? Is that even possible without making it too confusing (I mean, this is templated YAML, so it's already confusing 😄)
• ❓ I couldn't find where we specify the number of Zeebe partitions in the Optimize deployment - isn't that required? 🤔

The main blocker is that some golden file tests (for deployment for example) are pointing to the Operate chart, so the optimize deployment is not tested 😄

[npepinpe]

I think you'll need to regenerate the golden files, but after that, no blockers from side 👍

[Zelldon]

Thanks for your quick review @npepinpe

Regarding your comments:

clown_face Kind of weird that we can't use a single .helmignore for all the charts together man_shrugging

Yeah I can check this.

At this point, I feel like the secret helpers might be, uh, abstracted? Is that even possible without making it too confusing (I mean, this is templated YAML, so it's already confusing smile)

Yeah agree but tbh I don't want to rewrite the logic of bitnami, it is in general quite cool to use their helpers here. I could try, if I find the time, to add some abstractions around it.

question I couldn't find where we specify the number of Zeebe partitions in the Optimize deployment - isn't that required? thinking

As I wrote above this needs to be fixed with #286 and related discussion here https://camunda.slack.com/archives/C02UT3T8DQR/p1650536952860079 Please put your comments or opinions here if you have any :)

[RomanJRW]

Looking mostly at config this seems fine. Some minor comments and one probable missing config for the logout, but otherwise can't see any issue 👍

[RomanJRW]

Thanks for the heads up. Once JIRA wakes up, I'll create a ticket for us to change this in future 👍

[Zelldon]

Thanks @npepinpe and @RomanJRW for your time and review 👍 I applied all review hints and will now happily merge this :)