This message was deleted.

you'll need to export state an examine the error

pulumi stack export --stack foo > state.json

Thanks, @icy-controller-6092 for the response. So, ran that and the record that is claimed to be missing is in the generated JSON file:

{
    "urn": "urn:pulumi:stage1::infra::aws:lb:ApplicationLoadBalancer::lb-name",
    "custom": false,
    "type": "aws:lb:ApplicationLoadBalancer",
    "parent": "urn:pulumi:stage1::infra::pulumi:pulumi:Stack::infra-stage1",
    "sourcePosition": "project:///node_modules/@pulumi/lb/application.ts#54,9"
},

All the other records with a reference to the

ApplicationLoadBalancer

mentioned have the URN listed as a parent:

"parent": "urn:pulumi:stage1::infra::aws:lb:ApplicationLoadBalancer::lb-name",

It looks like the URN in your original error message is different

Sorry - I manually changed the actual name for confidentiality reasons.

Ah okay, just be careful when examining the state.json file because URNs must match exactly. I think there are some docs on how to manually edit stack state and then import it once done, have a read of those then start seeing what needs to be changed to fix it

Makes sense of course. In the original message it's complaining about a R53 resource (

urn:pulumi:stage1::infra::aws:lb:ApplicationLoadBalancer$aws:route53/record:Record::r53-record

), that depends upon a ALB resource (

urn:pulumi:stage1::infra::aws:lb:ApplicationLoadBalancer$aws:lb/loadBalancer:LoadBalancer::alb-name

). The R53 record (string) occurs twice in the JSON. Once as a "parent", the other as custom resource (

"custom": true

) in the JSON file. The ALB record string occurs 55 times in the JSON file, but every time as a dependency of another record (either a R53 record or a CloudWatch alarm - which is expected).

I wish I could help but I've never had much luck with the import process & I've already lost too many hairs from trying to get it work 😉

Interestingly though, the resource exists when I view the stack on the web/cloud.

when you say "resource exists" do you mean the actual AWS resource, or the resource entry in your stack's state? they are two seperate things

Well, the ACTUAL resource exists in AWS (it's active and is functional). AND If I go to app.pulumi.com and view my stack's resources, the "missing" ALB resource is in the list.

that is strange, if it's in app.pulumi.com then it should also appear in your stack export. Does your project have multiple stacks or just one?

Multiple stacks. It seems to be only this one stack that's having this issue.

• you have a resource named

r53-resource-name

and another named

lb-name

•

lb-name

is in your stack because you posted the snippet from state.json earlier, which is why you are also seeing it in web view • i assume the

r53-resource-name

is also in your stack's state - because the error message indicates it is there, but it referss to a missing resource

lb-name

• my best guess would be the URNs are not matching • did you perform any

pulumi up

actions using the

--target

flag?

• did you perform any

pulumi up

actions using the

--target

flag?

Actually, I did. BUT, it was on a DIFFERENT ALB (in the same stack).

# pulumi up --target 'alb-urn' --target-dependents

resource53

has a dependency on:

urn:pulumi:stage1::infra::aws:lb:ApplicationLoadBalancer$aws:lb/loadBalancer:LoadBalancer::lb-name

but in your snippet the URN for

lb-name

is:

urn:pulumi:stage1::infra::aws:lb:ApplicationLoadBalancer::lb-name

these URNs do not match, even though you mention you edited the URNs for privacy

I may have "mis-grabbed" earlier (adding to the confusion, so I apologise for that).

I have a feeling when you ran with

--target

it updated

lb-name

- but not the reference in

r53-resource-name

- which broke the reference you can go through your history on the Updates tab in Pulumi Cloud to see what each update has applied.

Scanning the resources and history, it's like this

$aws

resource never existed...