After I upgraded to Pulumi 3.0, and using `pulumi_...
# aws
f
After I upgraded to Pulumi 3.0, and using
pulumi_aws
4.0.0, it looks like the DynamoDB Table resource no longer has
hash_key
as an Output. Has anyone else run into this?
b
Are you using Python SDK @full-artist-27215?
f
Yes, sorry... I am using Python
b
it's definitely still in our SDK
f
It doesn't appear to be returning anything now, though.
b
can you open an issue on pulumi-aws repo and have a small snippet that shows the behaviour? I want to try and replicate this ad what you have found
f
Yup, I'll work up a minimal test case. Thanks!
🤔 This is very bizarre... having trouble replicating with a minimal testcase, despite earlier testing. I'll do some further digging; consider this a non-issue in the meantime. Sorry for the trouble!
The plot thickens... I have created a subclass of a DynamoDB Table resource that specializes it (ensuring all instances have PAY_PER_REQUEST billing, etc.). It is a direct subclass, and not a Component Resource. Before Pulumi 3.0, the inputs of the base DynamoDB Table resource were available as outputs of my subclass. Now, it seems that the inputs are no longer available, but the explicitly defined outputs of the base DynamoDB Table resource (e.g.,
arn
,
id
) are available on my subclass. Does that sound right? Should we not create subclasses of Pulumi resources, or is there a way to do that that exposes the inputs as outputs?
r
f
@red-match-15116 @broad-dog-22463 Sorry for the delay; here's a reproduction: https://github.com/christophermaier/pulumi-3.0-subclass-bug
Looks like someone else has run into a similar situation: https://pulumi-community.slack.com/archives/CDE799L1M/p1619056647129600
m
Thanks for putting together the repro, @full-artist-27215! Mind opening an issue at https://github.com/pulumi/pulumi/issues with a link to your repro?
f
@microscopic-pilot-97530 sure thing
m
Thank you! We’ll take a look.
🙇 1