My role policy is unable to create new roles. Is there a way to set default function props / stack synthesizer for the debug stack, so I can provide it the necessary roles and doesn’t try to create them?

Hi guys, is it possible to set

stage

dynamically in SST? For example in my case based on git branch. In Serverless Framework I’ve accomplished this via javascript file variables. Inside the file was a function which extracted git branch and returned it as a string which then became value of the

stage

variable.

I have an HTTP API that uses a custom domain as such :-

customDomain: {
    domainName: `${stageSubDomain}.${process.env.HOSTED_ZONE_NAME}`,
    hostedZone: hostedZone,
    path: process.env.API_VERSION, // this is 'v2'
  }

I’m now trying to create another HTTP API in the same stack with the custom domain path set to ‘v3’ for versioning sake but this fails and errors out with this message :-

<http://abc.pqr.xyz|abc.pqr.xyz> already exists in stack arn:aws:cloudformation:region:xxx:stack/dev-api-v2-media-stack/xxxxxx

Is there a way to achieve path level versioning with route53 custom domains with HTTP APIs?

Hey does anyone know what type I should be using for my API gateway events if i am using cognito identity pool auth like seen in the guide? So far have been using

APIGatewayProxyEventV2

from

@types/aws-lambda

fine, however have been trying to get the identity pool id like seen in the guide using

event.requestContext.authorizer.iam.cognitoIdentity.identityId

however the type seems to not think this field exists (it definitely does). My version is

"@types/aws-lambda": "^8.10.92",

so might be related to that, but figure someone might have some insight into correct version to use if so 🙏

Hey all, Is there anywhere I may be able to find an example of using ApiGatewayV1Api with request schema validation json file input. I get from using dasm, you'd use the L0 apigateway.CfnModel & apigateway.CfnRequestValidator somehow, but I can't see where they would be added into ApiGatewayV1Api

[Opinions Wanted]: What might be the pros/cons of using one of these two solutions over the other for keeping s3 buckets insync? 1. Using AWS Step Functions 2. Using Simple AWS Lambda Particularly interested in major differences in: • features • cost • stability • speed

On https://serverless-stack.com/chapters/call-the-list-api.html there's a line

return API.get("notes", "/notes")

. What's the first argument? "notes"? What does it match? I'm getting

API notes does not exist

in my app (which is a bit different).

Is there a difference between an edge function and a function? Like do I need to import

import { EdgeFunction } from "@serverless-stack/resources";

or something? Or is what makes it “edge” that I’m doing some config work with

cfDistrobution

Hello everyone... I'm trying to get my application to deploy via Seed, but I'm running into some errors that I'm not getting when I start, build and deploy locally. When I try to deploy, for my StaticSite, I get this error when it tries to install the npm deps:

$ cd /tmp/seed/source
INFO: Running before_build hook
$ echo "Before build"
0.08 s
$ cd website && npm install
82.19 s
$ npx sst build --stage dev --region us-west-2 --verbose
0.60 s
npm ERR! could not determine executable to run
npm ERR! A complete log of this run can be found in:
npm ERR!     /root/.npm/_logs/2022-02-16T19_55_37_431Z-debug.log
ERROR: There was an error synthesizing your app.

I have a build spec:

before_build:
 - cd website && npm install

I feel like there's something simple I'm missing, but I can't figure it out. Thanks for any help!

[Possible Bug]:

Received response status [FAILED] from custom resource. Message returned: Waiter InvalidationCompleted failed: Max attempts exceeded

This is the second time I’ve gotten this today. No rhyme or reason though, as consecutive tries seem to work.

How is packaging/transpilation being done in SST? And how source maps are handled there? My case: I have lerna-based repository and I want to import some local lerna TypeScript package (

@package/logger

) to my function. - that works ✅ When error is thrown inside the local lerna package (

@package/logger

), the stack trace is broken and doesn’t show much. ❌ For example:

{
  "errorType": "some-error",
  "errorMessage": "hello",
  "name": "some-error",
  "stack": [
    "some-error: hello",
    "    at l (/var/task/src/lambda.js:1:386)",
    "    at Runtime.p [as handler] (/var/task/src/lambda.js:1:467)",
    "    at Runtime.handleOnce (/var/runtime/Runtime.js:66:25)"
  ]
}

The stack trace should lead to

@package/logger

files.

Hey all! We had to specify the naming "Qualifier" for the bootstrapping stack (CDKToolKit) due to restrictions on AWS resource-naming imposed by our IT dept, which is fine for normal deployment if we override the

DefaultStackSynthesizer

. However, when we try to deploy a debug stack for local debugging with

sst start

, it gets caught in a failing loop in which it keeps trying to create a new bootstrapping stack, which fails because the default resource names violate our restrictions. How can we force sst/cdk to use our already-existing bootstrapping stack?

the command seems to deploy 2 stacks, im not sure is this is intended or i misconfigured my backend.

BlackFenix2-bip-debug-stack and BlackFenix2-bip-bip-stack

Do we have a example of a stack with auth and AWS Amplify?

** this is not a SST issue but CDK within SST I’m a bit in over my head and looking for some help. My goal is to have a function run at the edge, every time a resource is called from my static site. The function is located with the rest of the my functions (in

src/

. my project is in

us-east-1

. Below is a simplified version of my code.

import * as sst from "@serverless-stack/resources";

import { LambdaEdgeEventType, experimental } from "aws-cdk-lib/aws-cloudfront";
import * as lambda from "aws-cdk-lib/aws-lambda";

export default class MarketingStack extends sst.Stack {
  constructor(scope, id, props) {
    super(scope, id, props);

    const originRewriteHandler = new experimental.EdgeFunction(
      this,
      "OriginRewriteFunc",
      {
        runtime: lambda.Runtime.NODEJS_14_X,
        handler: "src/rewriter.main",
      }
    );

    const staticSite = new sst.StaticSite(this, "Site", {
      path: "website",
      errorPage: "404.html",
      buildOutput: "public",
      buildCommand: "npm run build",
    });

    staticSite.cfDistribution.addBehavior("/*", staticSite.s3Bucket, {
      edgeLambdas: [
        {
          functionVersion: originRewriteHandler.currentVersion,
          eventType: LambdaEdgeEventType.ORIGIN_REQUEST,
        },
      ],
    });
  }
}

When I try to tun this I get a type error that it can’t read

bind of undefined

. Which makes sense since I have no idea what I’m doing 🤣 . Actually, I figure that it’s trying to connect the bucket and the lambda but I’m missing something. I don’t really kno how to debug this.

Is it possible that by enabling lambda source maps (like suggested here -

setDefaultFunctionProps

) the live VScode lambda debugging gets broken? When I remove that environment setting from the stack and deploy it, the debugger starts to catch the breakpoints. When I return it back and deploy it, it doesn’t catch any breakpoint. Is VScode lambda debugger be taking into considerations the source maps anyhow when handling the breakpoints?

Question: Your Cognito JWT examples use the CDK

Cognito

construct. Can the same be achieved with the

sst.Auth

construct? I was wondering why my User Pool wasn't showing up in the console when I realized this.

Anyone else encountering an issue in which you supply a

restApiName

to a new

ApiGatewayV1Api

, but it still ends up getting named

{stage}-{app}-{constructId}

?

does the sst.Auth class support adding social login via the userpool?

Howdy all, I have the unfortunate requirement of having to define a traditional (non Aurora) Postgres RDS instance, and then have an

Api

interact with it. Does anyone have an example of the CDK magic they performed to do the similar. I'm seeing wildly different examples out in the Internets, and I've come to trust the members of this community to have a much more precise and pragmatic approach to their CDK definitions. I'm only looking to have the RDS instance running for an actual deployment, for local I'm going to defer to a docker PG instance.

Hey all, I'm a beginner to sst and trying to create a sign up/login page where user will only put their phone number, and they will get an otp , after providing the otp the user will be logged in (no need of email and pass). I'm following the serverless stack to build up the web app in the sst and there is only email and pass are mentioned. Can it be done with mobile number in sst itself ? Please help me to solve this. Thanks

RDS is not in our region and we would like to use it. How should we proceed to change region? sst delete on each environment and then change the sst.json ? Are we missing something here?

How to handle the situation when you have a function that is an trigger event of a bucket and needs the name of the bucket as env? Is there a way to do this without a cyclic reference error? Part of the same stack, it works, but in a different stack it won’t work.

All of a sudden my

AWS_PROFILE=xyx sst start

isn't working as expected, and hangs after printing

==========================
 Starting Live Lambda Dev
==========================

SST Console: <https://console.serverless-stack.com/>...

the SST console site says

No stacks found for app deploy-controller and stage piers

- I'm not sure where it's getting this stage from as I've not started sst with that stage parameter, any suggestions on what I can do to get past this and be able to test the frontend locally again?

When i try to add an AWS cognito userpool to my serverless stack, i get this error:

TypeError: Cannot read properties of undefined (reading 'UserPool')
    at new BipStack (C:\Users\Asus\Documents\GitHub\bip\aws-backend\stacks\BipStack.ts:23:34)
    at Object.main (C:\Users\Asus\Documents\GitHub\bip\aws-backend\stacks\index.ts:17:3)
    at Object.<anonymous> (C:\Users\Asus\Documents\GitHub\bip\aws-backend\.build\run.js:94:16)
    at Module._compile (node:internal/modules/cjs/loader:1101:14)
    at Object.Module._extensions..js (node:internal/modules/cjs/loader:1153:10)
    at Module.load (node:internal/modules/cjs/loader:981:32)
    at Function.Module._load (node:internal/modules/cjs/loader:822:12)
    at Function.executeUserEntryPoint [as runMain] (node:internal/modules/run_main:81:12)
    at node:internal/main/run_main_module:17:47

Is there an example what information I need to pass from sst.RDS to my lambda to execute a query from?

anyone have any idea why the sst debugging becomes unavailable?

Hi, I'm following the guide on using Cognito for user authentication (https://serverless-stack.com/chapters/using-cognito-to-add-authentication-to-a-serverless-app.html) but I would like to avoid using AWS Amplify. One solution would be to log in users through an HTTP API and use, for example, a JWT to authenticate calls to API endpoints that I want to be accessible only to registered users. Unfortunately, however, I have not found any resources on how to manage Cognito users on the server side, any help would be appreciated, thanks.

Is there a way to attach a

DockerImageFunction

to a Bucket notification? I’m getting:

Type 'DockerImageFunction' is not assignable to type 'FunctionDefinition'.
  Type 'DockerImageFunction' is missing the following properties from type 'Function': _isLiveDevEnabled, localId, attachPermissions, getConstructMetadata ts(2322)

When I add a route to an

sst.WebSocketApi

that uses a Lambda integration, the source ARN of the API Gateway trigger that gets added to the Lambda function appears to be incorrectly formatted