airasia / terraform-google-gcs_bucket

Terraform module for storage bucket in GCP

Terraform module for storage bucket in GCP

Upgrade guide from v2.4.1 to v2.5.0

This upgrade will rename the load-balancer resource(s) for them to be coherently identifiable in the GCP console as Bucket LB resource(s) which helps for faster troubleshooting.

1. First, ensure that you have applied v2.4.1 of this module before proceeding.
2. Then upgrade gcs_bucket module version from 2.4.1 to 2.5.0
3. Run terraform plan - DO NOT APPLY this plan yet.
4. If your bucket module at v2.4.1 had var.create_bucket_lb = true (this value was automatically set to true for domain-named-buckets), then you maybe seeing several changes being planned by terraform here. Otherwise, you SHOULD NOT see any changes for your module in this plan. If you STILL see any changes being planned here, then please ensure that you have first reconciled and applied v2.4.1 properly before proceeding with the next steps.
5. Ensure that the changes being planned by terraform for v2.5.0 are ONLY for the following 6 resources and nothing more than that:
   1. google_compute_backend_bucket.bucket_backend[0]
      1. renames from backend-bucket-___ to bucket-backend-___"
   2. google_compute_global_address.lb_ip[0]
      1. renames from lb-ip-___ to bucket-lbip-___" - ⚠️ we will revisit this change in coming steps
   3. google_compute_global_forwarding_rule.fw_rule[0]
      1. changes ip_address - ⚠️ we will revisit this change in coming steps
      2. renames from forwarding-rule-___ to bucket-fwd-rule-___"
   4. google_compute_managed_ssl_certificate.mcrt[0]
      1. renames from cert-___ to bucket-cert-___"
   5. google_compute_target_https_proxy.https_proxy[0]
      1. renames from https-proxy-___ to bucket-proxy-___"
   6. google_compute_url_map.url_map[0]
      1. renames from lb-___ to bucket-lb-___"
6. Must not contain any other changes
   1. You SHOULD NOT see any other changes in this plan other than the 6 changes listed above. If you see any other changes, then please ensure that you have first reconciled and applied v2.4.1 before proceeding with the rest of the steps.
   2. In the next steps, we will revisit the IP address changes (marked with ⚠️ above) for backward-compatibility.
7. Use the module variable called lb_ip_name to override the name change of the lb_ip resource.
   1. NOTE: This ability is provided only for backward-compatibility so that existing systems that maybe referring to this IP address (eg: DNS etc) do not break due to changes in this IP address.
   2. RECOMMENDATION: If you do not have any such systems depending on this IP address (or if you have the means of updating the IP address value in those systems) then we recommend you to not use this variable at all. Instead, go ahead with applying the changes presented in this plan and then update the IP address values in those external systems with the new IP address produced here.
      1. You can find the current IP address of the load-balancer from the output attribute called lb_ip_address.
8. After deciding between 7.1 vs 7.2, run terraform plan again and inspect the plans being proposed.
   1. If you see ANY OTHER CHANGES than the ones elaborated above, then please ensure that you have first reconciled and applied v2.4.1 and then repeat the above instructions before proceeding with the rest of the steps.
9. Run terraform apply
   1. The changes discussed above may take a maximum of 1 to 2 minutes.
10. If you decided to go with option 7.2 above, remember to update your external systems (if any) with the new IP address produced here.
    1. You can find the current IP address of the load-balancer from the output attribute called lb_ip_address.
11. Done!

Upgrade guide from v2.0.1 to v2.1.0

First make sure you've planned & applied v2.0.1. Then, upon upgrading from v2.0.1 to v2.1.0, you may (or may not) see a plan that destroys & creates an equal number of google_storage_bucket_iam_member resources. It is OK to apply these changes as it will only change the data-structure of these resources from an array to a hashmap. Note that, after you plan & apply these changes, you may (or may not) get a "Provider produced inconsistent result after apply" error. Just re-plan and re-apply and that would resolve the error.