DXE-4165 How best to handle long running Akamai Terraform jobs in CI/CD?

[glenthomas]

With some of the resources taking a long time to apply because of activation times, I am wondering how this is typically handled within automated CI/CD processes, which will often have a timeout less than an hour. Not only dealing with timeouts, but also being restricted on the rate of deployments, where changes must be batched and deployed infrequently because of long deployment times.

I have previously used AWS CloudFront, which has a wait_for_deployment setting. If enabled, the resource will wait for the distribution status to change from InProgress to Deployed. Setting this to false will skip the process. This allows the Terraform command to complete quickly, while the distribution gradually reaches all of the edge locations over a longer period of time.

There is no such option within the Akami resources, such as akamai_datastream, which can take over an hour to deploy.

Are there any strategies used for handling long-running Akamai Terraform deployments in automated workflows?

Terraform and Akamai Terraform Provider Versions

Terraform 1.9.4
Akamai Provider 6.3.0

Affected Resource(s)

• akamai_datastream

[lsadlon]

Hi @glenthomas,

Thanks for reporting it. We will investigate potential solution and go back to you. For now could you tell as if there are other resources that are problematic for you beside akamai_datastream?

BR,
Lukasz