v0.15.0-rc.1

A big focus this release was on fixing lingering issues with the Install PCF on GCP pipeline. Future releases will place a strong focus on each of the Install PCF pipelines for the other IaaSes.

Params changes:

• Install PCF on AWS

  • Remove TF_VAR_rds_db_username_ertsql
  • Remove TF_VAR_rds_db_password_ertsql

• Install PCF on GCP

  • Remove ert_sql_db_username
  • Remove ert_sql_db_password
  • Remove tile_upload_timeout
  • Change pcf_ert_ssl_cert to support 'generate' as the value (which is also the default)
  • Add ospman_uri

• Install PCF on vSphere

  • Add opsman_ssh_pasword

Pipeline changes:

• opsman_uri is no longer expected to be prepended with https protocol

• Install PCF on GCP

  • Add bootstrap-terraform-state job
  • Change where terraform state file is stored (root of bucket rather than in bucket/terraform)
  • Remove dependency on tfstate-version file
  • create-infrastructure now uses terraform state file properly
  • create-infrastructure no longer rolls back in the face of errors
  • configure-director no longer automatically runs after create-infrastructure
  • upload-ert is split out of deploy-ert so they can be re-ran independently
  • Reduce parallelism of create-infrastructure to ward off API errors

• Install PCF on vSphere

  • elastic-runtime is now checked less frequently for updates
  • elastic-runtime versions are now sorted by semver
  • elastic-runtime is now hardcoded to track 1.10.x

Task changes:

• upload-product-and-stemcell

  • Only try to upload stemcell when the product needs it
  • Skip downloading stemcell when it already exists in OpsMgr

• install-pcf-gcp/wipe-env

  • Add wipe of OpsMgr installation + BOSH director in addition to Terraform-managed resources

Bugfixes:

• Install PCF on GCP

  • Fix problem with incorrect OPSMAN_URI value passed in
  • Fix problem with incorrect value used for OpsMgr SSH password
  • Fix problem with SQL authorized users not being set correctly

• Install PCF on vSphere

  • Fix broken reference to pivnet cli

• czero/cflinuxfs2

  • Fix problem with the cleanup task that broke the aws CLI

• Remove reference to czero/busybox, which isn't a thing