This message was deleted.
# aws
s
This message was deleted.
s
What version of AWSX is in your
package.json
? Unless you specifically specified e.g.
npm i @pulumi/awsx@1.something
, you're using 0.4.x, which is TS-only and is not the version documented in the registry.
I recommend trying the 1.0 beta (and please please give us feedback on how it's working for you!)
v
1.0.0-beta.10
ok will give this one a try, thanks a lot! is it worth a note on the docs somewhere saying that the docs are for a non-stable version of the provider? cheers mate
s
So we do have a note on the registry page, but it may not have sufficiently caught your eye or the text may not have been clear enough. Anything obvious we could change to make it more clear?
v
ahhh ok, sorry thats my bad, i never normally look at the overview page, i normally head straight to api docs.
s
Not your bad - that's good usability feedback!
Can you upvote this issue? It's not exactly addressing your use case, but fixing the use case described would also give us the ability to add alerts above the tabs for any provider. https://github.com/pulumi/registry/issues/1356
e
@stocky-restaurant-98004 There's quite a bit of change to the 1.0.0 package version, and from my perspective it's for the better. Is it possible to say how stable the 1.0.0-beta10 is and how far away this is from a stable release? A clearer warning in the user guide docs on which main awsx version to use would also be nice:)
s
It may depend on the component. Are there any specific components are you looking at? I might be able to provide more context based on the component in question.
e
I'm currently building a ECS fargate service exposed via Application load balancer build into shared subnets. I wasn't able to use the 0.40.0 awsx version to deploy an ecs fargate service into two shared subnets since It uses awsx vpc recource, which I cant use in this scenario. Assumnig I haven't missed anything of coure
s
@elegant-architect-38580 If you're using the AWSX 1.0 beta VPC with the default options (which creates public and private subnets across 3 AZs), you're probably ok, but no guarantees since it's pre 1.0 after all. Based on the feedback I'm submitting as a result of using the AWSX 1.0 VPC for more advanced networking, I do expect some changes to the arguments.
e
I work in a platform team and have already deployed the VPC in a central network hub AWS account. Subnets are shared with the workload account where a product team creates the pulumi program deploying a fargate service. So the awsx resources for the reources in the workload account needs to input existing subnets to deploy resources .
s
Is the network hub defined in Pulumi?
e
The vpc’s are not
s
Actually, doesn't matter. I would recommend trying
getSubnets
to get your subnets. Filter by VPC ID (and maybe tag? depends on how you set up those subnets), e.g.:
Copy code
aws.ec2.get_subnets(
    filters=[
        aws.ec2.GetSubnetFilterArgs(
            name="tag:Name",
            values=[f"my-vpc-isolated1-*"],
        ),
        aws.ec2.GetSubnetFilterArgs(
            name="vpc-id",
            values=[YOUR_VPC_ID_HERE],
        ),
    ]
)
If you're sharing subnets among accounts, that VPC ID is gonna be stable so you can either hard-code it or add it as config to the stack if you have multiple environments.
e
Thanks, that is what I’m currently using. But I wasn’t able to combine it with the 0.40.0 awsx version of fargate and load balancer resources. From what I understand it only support awsx vpc as network input
s
Ohhhh, I see.
I would try 1.0 and see if that's any better for you. It should be. If it's not, please let me know or file a GH issue in pulumi-awsx so we can get some action on your feedback.
What language are you using BTW?
e
Typescript. I want to go for the 1.0 version but would of course feel more comfortable if the 1.0.0 had reached GA🙂