AWS RDS provisioning times out

[wilb]

I have a config block like:

resource "aws_db_instance" "test" {
    identifier = "test"
    db_subnet_group_name = "test"
    engine = "mysql"
    engine_version = "5.6.19"
    instance_class = "db.t2.small"
    multi_az  = true
    allocated_storage = "50"
    name = "test"
    username = "root"
    password = "password"
    vpc_security_group_ids  = [ "${aws_security_group.test-db-servers.id}", "${aws_security_group.test-app-servers.id}" ]
    backup_retention_period = 7
}

The provisioning will stall for a while (the instance creation does happen as expected) and eventually bail with:

aws_db_instance.test: Error: timeout while waiting for state to become 'available'
Error applying plan:

1 error(s) occurred:

 * timeout while waiting for state to become 'available'

Terraform does not automatically rollback in the face of errors.
Instead, your Terraform state file has been partially updated with
any resources that successfully completed. Please address the error
above and apply again to incrementally change your infrastructure.

I've tried creating this with publicly_accessible = "true" set but this doesn't help. The instance is created as expected, but in the AWS console it is still at the modifying stage for a good while after creation has timed out at the console. Is this just simply a need to increase the timeout,?

Obviously this leaves terraform in a state of inconsistency; I'm having to comment out my db instance provisioning after initial creation otherwise terraform then tries to recreate an already existing host.

[wilb]

Actually having just double checked the ultimate outcome what actually happens when you run terraform apply again is that it deletes the instance it's created (but thinks has failed):

aws_db_instance.test: Destruction complete
Apply complete! Resources: 0 added, 1 changed, 1 destroyed.

Interestingly even though this operation takes a long time too there is no timeout reported here.

Commenting out the host after provisioning doesn't help either as then terraform thinks I want to delete the host, resulting in it being destroyed anyway. So currently it's impossible to actually provision an RDS instance for me.

I'm using the latest codebase built this morning after #410 was fixed.

[pearkes]

Should be less frequent, increased the DB instance timeout to 40 minutes in 6f27f90

[geekmuse]

Still seeing this behavior even at the 40 minute threshold, particularly with SQL Server RDS instances

[JonRoma]

Seeing this with Terraform v0.7.2 and Oracle SE2 12.1.0.2.v4 RDS in a db.t2.micro instance type.

Times out right at 40 minutes.

[ossareh]

@pearkes it's not so much the timeout that is the issue, it's that any treatment to the RDS instance is then delayed until the next run. The timeout itself feels like it should be a variable; is that reasonable?

[mattsoftware]

I am seeing this too. Trying to restore an aurora instance with a large snapshot, takes over 40 minutes to do, but does complete. After the cluster is created, and terraform is re-run, it first destroys the cluster and then tries to create it again.

[ff-pspiess]

I am seeing this too, it takes more than 40 minutes for the cluster to come online. Is there a workaround?

[clearly]

+1

[jpriebe]

Seeing 40+ minutes to stand up RDS Aurora with a 30GB database deployed across two AZs. Please fix this issue; I'm not sure how I'm going to be able to use Terraform unless I either exclude RDS from the config or I fork terraform and compile my own version.

[rfink]

+1

[Hashfyre]

Seeing this with terraform v0.8.2, RDS, Postgres

Error applying plan:

1 error(s) occurred:

* aws_db_instance.test-postgres: timeout while waiting for state to become 'available' (last state: 'modifying', timeout: 40m0s)

Terraform does not automatically rollback in the face of errors.
Instead, your Terraform state file has been partially updated with
any resources that successfully completed. Please address the error
above and apply again to incrementally change your infrastructure.

[wilb]

I suspect no one is picking this up as it's a closed bug... Pinging @pearkes who upped the timeout originally?

[joseph-wortmann]

We just started having this with Postgres instances. DB created in AWS after about 15 minutes, TF waits for ~40 minutes and then times out. Rerun of TF shows no changes required.

[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 have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.