This message was deleted.

sparse-intern-71089

10/14/2020, 5:08 PM

This message was deleted.

✅ 1

elegant-island-39916

10/14/2020, 5:16 PM

To provide more context, the

pulumi.Output[List]

is a list of availability zones that were used in the creation of my AWS VPC in a different stack. They are being iterated over to create a

rds.ClusterInstance

for each availability zone

billowy-army-68599

10/14/2020, 5:20 PM

which language SDK are you using?

elegant-island-39916

10/14/2020, 5:22 PM

Python

elegant-island-39916

10/15/2020, 8:56 AM

Bumping this to see if anyone has a solution, I'm sure there's something trivial I'm missing. I've been unable to iterate through the output of another stack to create resources.

✅ 1

green-school-95910

10/15/2020, 1:16 PM

The reason for suggesting against creating resources inside

Output#apply

is that it causes incomplete graphs in some scenarios, like during the preview phase of the first deployment of that resource. But it is not forbidden. For your case I don't see any other option.

elegant-island-39916

10/15/2020, 1:22 PM

@green-school-95910 It didn't even create resources inside the apply function, I've got a nasty nasty hacky way of getting the availability zones now though:

Copy code

subprocess.run(["pulumi", "stack", "select", pulumi_config.require("network-stack")])
availability_zones = json.loads(subprocess.check_output(["pulumi", "stack", "output", "availability-zones"]).decode("utf-8"))

Surely this cannot be the only way of achieving this?

green-school-95910

10/15/2020, 1:23 PM

Have you tried just putting it inside apply?

elegant-island-39916

10/15/2020, 1:24 PM

I have, the resources I defined are not created

elegant-island-39916

10/15/2020, 1:25 PM

It looks something like this:

Copy code

def create_rds_cluster_instance(az_name):
    rds.ClusterInstance( ... )


network_stack = StackReference(pulumi_config.require("network-stack"))
availability_zones = network_stack.require_output("availability-zones")
availability_zones.apply(lambda zones: map(create_rds_cluster_instance, zones))

elegant-island-39916

10/15/2020, 1:38 PM

I've figured what's wrong, the actual creation should look something like this instead:

Copy code

def create_rds_cluster_instance(az_name):
    rds.ClusterInstance( ... )

network_stack = StackReference(pulumi_config.require("network-stack"))
availability_zones = network_stack.require_output("availability-zones")
availability_zones.apply(lambda zones: [create_rds_cluster_instance(z) for z in zones])

It's all working now. Cheers @green-school-95910!

elegant-island-39916

10/15/2020, 1:41 PM

Apologies for the brainfart, did not take into account

map

not evaluating all entries in the list

green-school-95910

10/15/2020, 1:51 PM

Yeah,

map

evaluates lazily as a generator.

👍 1

13 Views

Open in Slack

Previous Next

Pulumi Community

No matter how you like to participate in developer communities, Pulumi wants to meet you there. If you want to meet other Pulumi users to share use-cases and best practices, contribute code or documentation, see us at an event, or just tell a story about something cool you did with Pulumi, you are part of our community.