Hi - just trying out the simple hello world and I'm getting

Stack 'arn-aws-iam--xxxxxxxxxxx-my-stack' does not exist

OK, I figured it out by looking in the sst-debug.log and a little more googling. Apparently I needed to go into CloudFormation and delete the CDKToolkit Stack that was in there.

I have a pretty straightforward stack definition to try out a WebSocket API but, when running

sst start

it deploys and starts up but then logs the message

Missing region in config

and aborts with

Process finished with exit code 130 (interrupted by signal 2: SIGINT)

The deploy command I’m using is

sst start --stage ross-dev --region eu-west-2

The stack itself looks like this:

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

export default class WebSocketStack extends sst.Stack {

  private api: sst.WebSocketApi;
  private connectionsTable: sst.Table;

  constructor(scope: <http://sst.App|sst.App>, id: string, props?: sst.StackProps) {
    super(scope, id, props)

    this.connectionsTable = this.createConnectionsTable();
    this.api = this.createApi();
    this.api.attachPermissions([this.connectionsTable]);

    this.addOutputs({
      ApiEndpoint: this.api.url
    })
  }

  private createConnectionsTable(): sst.Table {
    return new sst.Table(this, 'connections-table', {
      fields: {
        PK: sst.TableFieldType.STRING,
        connectionId: sst.TableFieldType.STRING,
      },
      primaryIndex: {
        partitionKey: 'PK',
      }
    })
  }

  private createApi(): sst.WebSocketApi {
    return new sst.WebSocketApi(this, 'WebsocketApi', {
      accessLog: false,
      routes: {
        $connect: "src/main/handlers/wsHandlers.webSocketHandler",
        $disconnect: "src/main/handlers/wsHandlers.webSocketHandler",
        sendMessage: "src/main/handlers/wsHandlers.webSocketHandler",
      },
      defaultFunctionProps: {
        environment: {
          CONNECTIONS_TABLE: this.connectionsTable.tableName,
          REGION: this.region,
        },
      }
    })
  }
}

Any thoughts on what could be going on?

I'm starting a new project (

npx create-serverless-stack@latest notes --language typescript

), then, after running

npm run start

I'm getting

Missing region in config

. Does someone know how to fix this?

In my SST app, there is a “core bucket” which keep: configuration files, email templates and images. To prevent any manual operation after first deploy, I decided to keep in my monorepo a folder called “assets” which contains a “core-bucket” folder which inside are all files I want to publish on S3. To keep sync this folder and bucket I tried to us S3Deployment cdk construct

new s3deploy.BucketDeployment(this, 'core-bucket-deployment', {
      sources: [s3deploy.Source.asset('./backend/assets/core-bucket')],
      destinationBucket: this.core_bucket.s3Bucket,
});

It works nice but I have a problem about permission, because some of these file (example app-logo.png) I want it with public permission but I can’t find a way to set permission on file. Probably S3Deployment is not a right way to solve my need. Do you know some solution? Actually I think the only solution is to use aws-sdk after cdk bucket creation, to manually set permission on files loaded with S3Deployment, or directly upload that files through sdk.

Did anyone get PlanetScale + Prisma (or just Prisma) to work with SST?

Hello! Any reason why the AppSync construct does not contain a domainName attribute? It's declared on the CDK but can't find it anywhere on the SST construct

Hiya folks, I'm playing with the Cognito JWT Authroization example, but trying to use an AuthStack to create the cognito user pool for JWT based authentication. Can make it work, with two caveats 1) I need to manually enable the ALLOW_USER_PASSWORD_AUTH Authentication flow through the cognito console, and 2) Less of a caveat, more of an efficiency thing, the Auth stack creates an Identity Pool for IAM authorization by default, which I'd like to prevent / turn off. Appreciate any thoughts on achieving either of this ends! 🙂

hey Guys, i updated sst to

0.69.5

but i’m still getting the error addressed in that release when running

sst start

https://serverless-stack.com/examples/how-to-use-datadog-to-monitor-your-serverless-app.html Check this example for V2 usage of datadog.

can anyone help with translating this to SST 1.0? (within an

sst.Api

definition)

defaultAuthorizer: new HttpUserPoolAuthorizer(
        'HttpUserPoolAuthorizer',
        props.cognitoAuth.cognitoUserPool!,
        {
          userPoolClients: [props.cognitoAuth.cognitoUserPoolClient!],
        }
      ),

I’m not sure how to get the issuer and audience from an

sst.Auth

instance

Hello, I want to use NextjsSite construct with Cognito for a demo. What do you use for Cognito auth in Next.js?

Anyone here could help in providing an example for writing test cases for SQS SendMessage?

migrating to 1.0 won’t result in any deleting & replacing of resources, will it? sometimes CloudFormation wants to delete/replace something just because a resource name changes. I’m paranoid about it messing up something in production

I’ll just start a migration thread here, since I’m sure I’ll have more questions.. 🙂 starting with this: I can’t get

sst.Api

customDomain working:

Type '{ cdk: { domainName: string; certificate: acm.ICertificate; }; }' is not assignable to type 'string | ApiDomainProps | undefined'.
      The types of 'cdk.domainName' are incompatible between these types.
        Type 'string' is not assignable to type 'IDomainName'.ts(2345)

the documentation says domainName can be a string, yet it’s complaining about that

could someone help me understand the new

cdk

property in 1.0? like when defining a new EventBus, there’s

rules.myRule.pattern

and

rules.myRule.cdk.rule.eventPattern

. why are there two ways to define a pattern?

is there a way to deploy a stack without having all the Lambda functions build first? like if I was only doing an infrastructure change

hey guys I’d like to get some help with AWS organizations I have an account that I use for my personal projects, and I started using AWS Organizations/SSO to kill all of the IAM users I used to have, but now I can’t see the billing dashboard from the accounts logged in through SSO - even though I gave full access to the AWS account and to the billing dashboard. does it require any additional steps?

Is there a way I can execute this command in parallel (so I don't need to wait server to start before start my frontend)?

"start": "sst start && cd src/frontend && yarn dev",

Is there a way to change or adjust the behavior of

esbuild

in sst. Specifically, how does one use

--externals

to control bundling behavior? If not, is the recommended path to run

sst build/start ..

on an already transpiled src use the customized

esbuild

process.

When trying to use the

HttpUserPoolAuthorizer

from

@aws-cdk/aws-apigatewayv2-authorizers

to create a Cognito jwt authoriser I get the following error: Is this a version incompatibility issue between SST and Cognito?

Error: construct does not have an associated node. All constructs must extend the "Construct" base class
    at Function.of (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/core/node_modules/constructs/src/construct.ts:30:13)
    at new Node (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/core/node_modules/constructs/src/construct.ts:75:12)
    at new ConstructNode (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/core/lib/construct-compat.ts:291:24)
    at Object.createNode (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/core/lib/construct-compat.ts:78:11)
    at new Construct (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/core/node_modules/constructs/src/construct.ts:575:26)
    at new Construct (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/core/lib/construct-compat.ts:75:5)
    at new Resource (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/core/lib/resource.ts:151:5)
    at new HttpAuthorizer (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/aws-apigatewayv2/lib/http/authorizer.ts:159:5)
    at HttpUserPoolAuthorizer.bind (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/aws-apigatewayv2-authorizers/lib/http/user-pool.ts:59:25)
    at new HttpRoute (/home/robert/code/HeyDaily/heydaily-serverless/node_modules/@aws-cdk/aws-apigatewayv2-alpha/lib/http/route.ts:196:45)

Hi all! Currently I'm evaluating SST in combination with Seed.run for customer's project. The project is a completely serverless app consisting of a bunch of lambda function, step functions, api gateway endpoints and some other resources. The lambda functions are either written in "plain" python or python in docker images (containerized lambdas) due to high memory requirements when working with AI models. My question is about unit testing - the python source code including the Pipfile is in the src/<service_name> subfolder. When enabling unit tests in the seed pipeline it only checks for a Pipfile in the project root-directory. At the moment I've added the python unit testing in the pre_deploy stage in the seed.yml. Is there any way to specify the subfolder of the service root dir where the Pipfile is located when configuring the unit tests in the pipeline?

Seems like my function defaults (vpc, layers) are not being propagated to my lambda functions created via

API

I set the defaults before instantiating the

APIs

. it's all one stack SST 0.69.5

Hi all! I'm trying to set an environment variable for a S3 bucket notification based on another stack and getting a cyclic reference. Is there a way to add the notifications after everything else has been created or another way of doing this? Thanks in advance

When I run

sst deploy

my function can talk to my RDS DB just fine When doing

sst start

... no dice, the function can't reach RDS anymore

Hi, any idea how to add a custom domain to cognito user pool ? If there is an example, it would be best.

Hello, I have created a queue and attached to lambda function but somehow the consumer lambda is invoking twice though I was manually deleting the message after processing. Could any one help me with this?

Hi team, I am not able to edit custom domains on Seed despite removing custom domain config from my code. I have removed any route 53 records too. But the Seed build log outputs always shows the custom domain while my local build does not. Any advice on this?

Hi 👋 I see that a number of people have used Typeorm with SST. When using Typeorm we are seeing lambda cold start times of ~3.5s. Has anyone else experienced this issue or have any suggestions?