Improve UX around Azure params

[zmb3]

Just went through my first deploy with the Azure install pipeline and a few things could have made the experience better:

1. The comments in the params file say that if azure_access_name is empty, ERT will use internal storage. Then for each of the storage containers they say Required if Azure blob storage is used as external storage by ERT, which is determined by azure_access_name being not null. There are a few problems here:

• there is no azure_access_name param - I believe it was renamed to azure_account_name.
• I attempted to install using internal storage (by leaving azure_account_name empty, and got an error that the field cannot be empty. It seems this use case is broken.

2. The readme says:

Once DNS is resolvable, run the config-opsman-auth job. From there the pipeline should automatically run through to the end.

We didn't find that to be the case and had to manually trigger the next job. Are we missing a trigger: true somewhere?

3. We had to reverse engineer the proper value for ha_proxy_lb_name by looking through the tasks and the terraform files. The correct value for this property is always {{azure_terraform_prefix}}-web-lb. Can we remove this property and compute it automatically so as not to introduce another potential source of error?

4. I noticed that the terraform creates LBs for HAproxy, MySQL proxy, Diego Brain SSH, and TCP routing, but only configures the LB for the HAproxy job. If we're going through the trouble to create this infra shouldn't we configure it for the other jobs as well?

[cf-gitbot]

We have created an issue in Pivotal Tracker to manage this. Unfortunately, the Pivotal Tracker project is private so you may be unable to view the contents of the story.

The labels on this github issue will be updated when the story is started.

[abbyachau]

Hey @zmb3 thanks for this feedback. Can you let us know what version of the pipelines you were using? Thanks.

[zmb3]

We were on commit c5bc203 on master.

[abbyachau]

Hi @zmb3, thanks for the feedback.

I've broken up your feedback via tracker stories so that they could be more easily consumed by engineers:

1. azure_access_name issues - the comment has been updated. The story to look into the empty string issue can be found here: https://www.pivotaltracker.com/story/show/154223925

2. https://www.pivotaltracker.com/story/show/154224540

3. https://www.pivotaltracker.com/story/show/154234636

4. https://www.pivotaltracker.com/story/show/154256663

Please get in touch if you have any questions, thanks again.

[fredwangwang]

@pivotal-maya-kenedy @calebwashburn https://www.pivotaltracker.com/story/show/154224540
has already been completed. Can you review and prioritize the other concerns zmb3 mentioned?

[calebwashburn]

@pivotal-maya-kenedy Think we need to restore the stories for 1, 3, 4 as those got cleaned up inadvertently earlier.

[dtzar]

Our team has started working to resolve these issues here, but had a bunch of other bugs we found and had to fix first. A couple PRs are up there now (#364 #365) and more will come once those are merged.