Auto DevOps: dast_environment_deploy times out

Summary

The dast_environment_deploy job in the review stage of Auto DevOps appears to be timing out for test projects and customers when using full Auto DevOps with no overrides.

Steps to reproduce

1. Create a project on GitLab.com within a Gold subscribed namespace that's based on an application.
2. Create a Kubernetes cluster on GKE through Operations -> Kubernetes using the default options.
3. Install Tiller, Ingress, and GitLab Runner through GitLab Managed Apps.
4. Set a base domain, using nip.io if necessary.
5. Enable Auto DevOps.
6. Run a pipeline.
7. Observe the failure of the dast_environment_deploy job. Full error below.

Logs

 Application should be accessible at: http://dast-15529828-dast-default.35.204.11.84.nip.io
101 Waiting for deployment "dast-default" rollout to finish: 0 of 1 updated replicas are available...
105 ERROR: Job failed: execution took longer than 1h0m0s seconds

Example Project

• tristan/minimal-rails-app
  • Failed job

What is the current bug behavior?

dast_environment_deploy is timing out and failing.

What is the expected correct behavior?

The deployment should succeed.

Output of checks

This bug happens on GitLab.com: 12.5.0-pre cdba75bfa20

Environment info

• GKE Cluster 1.13.11-gke.14 (Master and nodes)
• GitLab Managed Apps used (Runner 12.5)
• Base domain: 35.204.11.84.nip.io

ZD: https://gitlab.zendesk.com/agent/tickets/137919 (GitLab Internal)