maintenance_window of google_sql_database_instance replica is stuck - not changing / updating / refreshing

[syedrakib]

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave +1 or me too comments, they generate extra noise for issue followers and do not help prioritize the request.
• If you are interested in working on this issue or have submitted a pull request, please leave a comment.
• If an issue is assigned to the modular-magician user, it is either in the process of being autogenerated, or is planned to be autogenerated soon. If an issue is assigned to a user, that user is claiming responsibility for the issue. If an issue is assigned to hashibot, a community member has claimed the issue already.

Terraform Version

• Terraform v0.12.29
• provider.google v3.41.0
• provider.google-beta v3.41.0

Affected Resource(s)

google_sql_database_instance

The issue:

My current TF state file shows the following for my Google CloudSQL Replica instance.

    maintenance_window: []

My new TF configuration for my Google CloudSQL Replica instance is

maintenance_window {
  day          = 1
  hour         = 23
  update_track = "canary"
}

✅ My TF Plan reads fine:

+ maintenance_window {
    + day          = 1
    + hour         = 23
    + update_track = "canary"
  }

Plan: 0 to add, 1 to change, 0 to destroy.

❌ I run terraform apply. I get the following error

Error: Error, failed to update instance settings for : googleapi: Error 400: Invalid request: Cannot update maintenance window for replica instances., invalid

❌ Meaning, the operation failed and nothing should've changed. But when i check my TF state file, i see that it has changed

"maintenance_window": [
  {
    "day": 1,
    "hour": 23,
    "update_track": "canary"
  }
]

❌ The state file shouldn't have changed since the terraform apply had failed.

Now, when i run the terraform plan, it produces the same unaltered plan as before.

+ maintenance_window {
    + day          = 1
    + hour         = 23
    + update_track = "canary"
  }

Plan: 0 to add, 1 to change, 0 to destroy.

---

❓ I am not able to move beyond this. My CloudSQL ReadReplica is constantly producing this same plan and I am constantly getting a 400 error when trying to terraform apply. Hence, i can't proceed.

I also tried to set the maintenance_window of the instance manually from the GCP console. However, GCP doesn't even provide any interface to modify maintenance_window for ReadReplica instances - which is why the 400 error in the first place.

---

Looking for help here. Using latest 0.12.x version of Terraform 0.12.29, and latest version of google provider 3.41.0

[edwardmedia]

@syedrakib can you post your config and the steps to reach the issue?

[syedrakib]

I used the v3.2.0 of the terraform-google-sql-db / mysql module for my work. It is these lines that are causing the issue.

Please note that, even though its providers.tf file says it uses terraform = "~> 0.12.6" and google = "~> 3.5", i am actually using the latest terraform 0.12.x version v0.12.29 and passing the latest google-provider version v3.41.0 to this module.

---

The fact that running terraform apply caused the state file to change even though it failed with a 400 error from google is the concerning part. That should be handled by the provider correctly, right?

[edwardmedia]

@syedrakib instead of module, can you run a single instance by removing unrelated field? Please provide the full debug log, if the problem still occurs.

[edwardmedia]

@syedrakib is this still an issue?

[edwardmedia]

@syedrakib closing this as no response received. Please retested by separating the provider code from the module. Feel free to reopen it if the issue can be reproed.

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. If you feel I made an error 🤖 🙉 , please reach out to my human friends 👉 hashibot-feedback@hashicorp.com. Thanks!