Hey all, curious if anyone has a good workflow for SST local dev and working with s3 files. I would like to have a bunch of JSON files in a local folder, read/write to them locally when doing local dev. Then when I deploy I would like those files to be sync'd to an s3 bucket and my read/write to work against s3. Maybe everything needs to be on s3, even during local dev? Any ideas on a clean way to work this?

Has anyone ever experienced inconsistent deploys between CI and local deploy? I’m deploying a fargate service and deploys from my local work fine, but the exact same commit is stuck in UPDATE_IN_PROGRESS for the ecs service (i let it time out once, 3hrs I think it took)… the weird thing is that this suddenly happened and I can’t find a cause for such a problem.

I try to deploy frontend stack but get

Invalid request provided: One or more of the CNAMEs you provided are already associated with a different resource.

where should I go and what should I delete?

Hi everyone, I am trying to add redis to my stack. Is there a way to get the VPC and subnets that were created and use them for redis. I looked around and cannot find a way to get vpc. Thanks

Hey everyone, Been trying to work with AppSync on SST, ran into an issue where in the SST console - My GraphiQL interface doesn't seem to load via SST console in the browser or via directly visiting my https://<aws appsycn generatedurl>/graphql endpoint output by my stack. See attached screenshot of the SST console. Literally upon load it reports the error and also won't load any schema info

"error": "Unexpected token < in JSON at position 0"

The 2nd screenshot is attempting to load the interface within the browser. The only way I can interact with the GraphQL api is via the AWS console "AppSync" console where it works fine, loads the schema, I can perform queries & mutations etc. As well as use the Appsync console to authenticate users Any one else ran into anything similar? ------ UPDATE ------ Ok quite a silly mistake on my part. I figured out how to load the UI for GraphQL in SST Console - I had to manually put my output of the GraphQL URL into SST Console, and additionally add a HTTP header to my request.

{
  "x-api-key": "input output API key Here"
}

Hey guys, My “sst build” command is not showing me these kind of errors. It builds the project correctly, and doesn’t fail. I would like to have a failed build in case the contracts are not met. Is there a way that I can setup it?

Hi 👋 I try todo

auth

as as in the guide but with

nextjs

and I get an issue when I use Amplify on the frontend and try to deploy the stack.

Error: Command failed with exit code 1: ./node_modules/.bin/next build
warn  - The `target` config is deprecated and will be removed in a future version.
See more info here <https://nextjs.org/docs/messages/deprecated-target-config>
warn  - Compiled with warnings

• I use all the latest packages in my package.json • I have ssr:true in Amplify config • It builds fine if I run build locally for the frontend • It deploys fine if I remove Amplify config from the frontend I understand this might not be related to sst but maybe someone had this issue and somehow resolved?

Hello everyone, Does anyone know why the permissions might not be attaching to the lambda function set at this route Im trying to extend one of the examples and have a lambda which I would like to return a list of all the users in the user pool, however, the iam policy isn't applied to the api route

// Create an HTTP API
    const api = new sst.Api(this, "Api", {
      // Secure it with IAM Auth
      defaultAuthorizationType: sst.ApiAuthorizationType.AWS_IAM,
      routes: {
        "GET /private": "src/private.handler",
        // Make an endpoint public
        "GET /public": {
          function: "src/public.handler",
          authorizationType: sst.ApiAuthorizationType.NONE,
        },
        "GET /cognito": {
          function: { 
            srcPath: "src/",
            handler: "getCognito.handler",
            environment: {
              "userPoolId": auth.cognitoUserPool?.userPoolId ?? ""
            },
          },
          authorizationType: sst.ApiAuthorizationType.NONE,
        } 
      },
    });

    api.attachPermissionsToRoute("GET /cognito", [
      new iam.PolicyStatement({
        actions: ["cognito-idp:ListUsers"],
        effect: iam.Effect.ALLOW,
        resources: [
          `arn:aws:cognito-idp:${this.region}:${this.account}:userpool:${auth.cognitoUserPool?.userPoolId}/*`,
        ],
      }),
    ])

Hey team I'm having a little trouble with debugging my graphql lambda. I'm trying to implement typegraphql and getting some errors. The problem is, I can only see these errors on 'production'

npx sst deploy --stage prod

. I cannot see errors on the debugging stack

npx sst start

. When I start the debugging stack and query my graphql lambda all I can see from the browser is

{"message":"Internal Server Error"}

and all I see from the command line is

8e29906c-8622-4ab7-98fd-5bbcb2c97764 REQUEST dummy-graphql-apollo-my-s-ApolloApiLambdaGET95C283-IaPKrXb9HmmO [src/lambda.handler] invoked by API GET /

and eventually I'll see this (looks like the call to the lambda just times out)

e30f62ab-2b81-4c6d-8fea-80baf75a8846-1651498094641 Failed to send a response because the Lambda Function timed out. If this happens again, you can increase the function timeout or use the --increase-timeout option with "sst start". Read more about the option here: <https://docs.serverless-stack.com/packages/cli#options>

However if I deploy this to prod I see a real error, I can paste in thread. This is a small git repo that should replicate the issue https://github.com/unwrap-nlp/graphql-apollo Do you mind helping me figure out how to see errors in the debugging stack?

Does anyone have examples of storing

env

variables in SSM? How are these initially loaded? Do you put all SSM params in one stack or is this integrated into your other stacks? Do you load them from env var file and then set them or set them in the UI?

Good afternoon everyone, I have a backend that I need to duplicate. Normally I would install the serverless lib in my terminal, and run a

serverless deploy

, and my functions would be pushed to lambda in my aws account. But I'm checking, there were big changes. Is it possible for me to deploy my api like I did before? at least initially for me to plan for migration?

Has anyone had recent luck with getting typeORM implemented with SST? I've tried previously mentioned solutions for externalModules/nodeModules for pg-native but cannot seem to by-pass the error when running sst build/seed deploy:

Error: There was a problem transpiling the Lambda handler: > node_modules/pg/lib/native/client.js:4:21: error: Could not resolve "pg-native" (mark it as external to exclude it from the bundle, or surround it with try/catch to handle the failure at run-time)
    4 │ var Native = require('pg-native')

Note - The above works locally with sst start I've also tried an alternative solution which involves having a mocked pg-native library (which just returns an empty module). This actually works with sst build & seed deploy, BUT then sst start runs into issues running typeORM locally. See attached error. I'm currently at a stand still of either typeORM works locally OR online, but have been unable to find a solution that works for both.

👋 Did anyone successfully completed the tutorial with

nextjs

instead of

create-react-app

? I get issues with deployment with auth (Amplify). Locally things works fine. Here's my previous thread: https://serverless-stack.slack.com/archives/C01JG3B20RY/p1651472910967679

Does anyone know where the

.dockerignore

should be placed to be picked up by the SST/CDK build of a docker lambda function?

Anyone know if these route53.HttpsRedirect preserve the path on a url? Eg.

<https://domain.com/blah>

to

<https://www.domain.com/blah>

Serverless-framework expects the values at

vpc.securityGroupIds

&

vpv.subnetIds

to be Arrays. The easier solution is just wrap your

${ssm:...}

calls in `[]`'s So for example:

vpc:
  securityGroupIds: [${ssm:terraform-output/$self:provider.stage}/vpc/security-groups/default/id/value}]
  subnetIds: [${ssm:terraform-output/$self:provider.stage}/vpc/subnets/private/ids/value}]

Anyone add

constructs

as a dev dependency and then have type issues on

stack

when used on SST components?

hey guys! Sorry if this is off topic, but hoping someone can point me in the right direction 🙏 I want to create API Keys to allow customers to use our API. I’ve set that up and it was super easy 👍 but what is best practice for how to identify the customer by their API key?

Anyone else having issues with debugging in vscode, debug panel buttons become unresponsive and requires vscode to be killed.

Morning guys! I have a service which is 100% handle by Seed. I am trying to add a new service, Seed see them both and my pipeline is successful, but Seed does not seem to see my two new endpoints so I can configure my domain on them. Anyone has any idea if I have to do anything special for Seed to see them? Pretty sure I never had to do anything special before. I was just going into

Settings / Edit Custom Domains

Thank you in advance!

Hello team, I have added my domain to Route 53 in a hosted zone. For testing I have deployed once in testing and it worked but from next deploy getting this error.

It seems you are configuring custom domains for you URL. And SST is not able to find the hosted zone "<http://app-serverless-cli.text-shot.com|app-serverless-cli.text-shot.com>" in your AWS Route 53 account. Please double check and make sure the zone exists, or pass in a different zone.

I tried deleting

npx sst remove — stage  {my stage name}

but still same issues.

I'm trying out the new functional stack approach with SST 1.0.0 and could use a pointer on sharing resources across stacks.

Should I be able to deploy a single stack by name using functional stacks? For example, if I have

app.stack(StackA).stack(StackB).stack(StackC)

, should I be able to deploy stackB to a

shared

stage?

yarn run sst deploy stackB --stage=shared

Based on the error message I'm receiving, I cannot:

Stack shared-my-service-stackname is not found in your app.

I know I can deploy the entire app to the

shared

stage, but that's not what I want here

Hi @thdxr, I am trying to implement https://github.com/serverless-stack/kysely-data-api and getting the following error. I can confirm that the module does exist in node_modules. Thanks

error - Error [ERR_MODULE_NOT_FOUND]: Cannot find module '/Users/lgupta/Documents/Projects/checkout-com-payments-app/node_modules/kysely-data-api/dist/esm/data-api-driver' imported from /Users/lgupta/Documents/Projects/checkout-com-payments-app/node_modules/kysely-data-api/dist/esm/data-api-dialect.js

v1 upgrade: • I can't find a replacement for CORS

allowMethods: [CorsHttpMethod.ANY],

. A value of

["*"]

should (less clearly) do it, but that isn't an accepted value. I see SST is using CorsHttpMethod internally, but won't let me.

v1 upgrade: • Is the lack of support for

events.CronOptions

intentional for

sst.Cron.schedule

? Seems like a huge loss for the only composable way to create a schedule to be gone.

v1 upgrade: •

PolicyStatement

not an acceptable type in

Permissions

?

Anyone has a link to a good guide if I want to transform all the API in my serverless.yml into some SST stacks? Most of my endpoints are authorised with Cognito.

has anyone successfully deployed an open source CMS (something like contentful or webiny) with SST?

Congratz on 1.0!!! I started with 0.39 on a small company project. Now I plan to make it the core of our tech stack for all new projects. The .cdk prop on the constructs is more than welcomed 😄 The upgrade to

1.0.2

went fine up to this error that I cannot wrap my head around, used to work fine on

0.65.5

node_modules/aws-cdk/lib/diff.d.ts:1:26 - error TS2307: Cannot find module '@aws-cdk/cloudformation-diff' or its corresponding type declarations.

1 import * as cfnDiff from '@aws-cdk/cloudformation-diff';

The current fix is to install the dependency myself:

"@aws-cdk/cloudformation-diff": "2.23.0"

, is this a bug on aws side?