Hello, I want to restrict access to an api gateway for only a specific other lambda function. is there a way to do that?

anyone know how to properly kill hung processes w/ SST?

? Something is already running on port 12577.

I did

kill -9

for any related processes. doesn’t seem to do the trick

Is there a way to setup API gateway to handle trailing

/

gracefully e.x

abc/hello/{123}

and

abc/hello/{123}/

should be both accepted and go to the same function

Hello folks, so I have secured my API as per https://serverless-stack.com/examples/how-to-add-cognito-authentication-to-a-serverless-api.html and tested everything works ok using

npx aws-api-gateway-cli-test

Now my question is do I make authenticated calls to the api from my client browser. I do not want to use Amplify. So far I have managed to login and fetch tokens using amazon-cognito-identity-js which is all very nice and gives me access+id+refresh tokens but once I have that how do I wrap things up to properly call the /private entrypoint in my API ? Thanks !

I love using the AWS CDK and for my own project, I added in the concepts of stages so I can can have

dev

,

test

,

prod

and so on. I stumbled on serverless-stack and realised where SST was better than my own (hastily) home rolled solution and I've looked into porting. I have one question regarding the management of stages/deploying. In my current CDK project, I have multiple "apps" (that is

bin/*.ts

) to represent different logical areas of infrastructure. For example a "global" app for things like Hosted Zones, an "identity" app for different identity concerns, an "api" app for deploying my public facing API into multiple regions (I'm using Regional API Gateways with Route 53 latency policies). My CI/CD pipelines can then run each app as required into the correct regions/stages and it all works. However with SST there are no more "apps", just the

stacks/index.js

. I'm wondering how to port my CDK apps to the SST setup to maintain my existing separation of stacks into their logical areas.

Is it possible to define a

ApiGatewayV1ApiRouteProps

instance without a Function, but just an API Gateway Mock Integration? Or will I get errors from SST?

Can you have multi path sgement keys in SST? eg:

/accounts/{id}/transactions

? All the examples are single path segments like

/notes

but the app I want to port to SST have multi path

Does anyone now how to configure Babel for an app based on the sample app? I’ve fully borked my testing setup for my example app. When I run

npm run test

at the project root i get

cannot find module babel-jest

with a project path pointing to:

node_modules/@serverless-stack/cli/scripts/util/babelJestTransform.js

if I install babel and drop in a config

babelJest.createTransformer is not a function

also in the CLI. How i got here was I added a test in the react app, then deploy’s broke on

Seed.run

see the previous thread here. At that time, at least the tests worked for the App, just not the react app. https://serverless-stack.slack.com/archives/C01JG3B20RY/p1633213092119300

I've got a weird interaction. For some of my services, I'm getting

{"message":"Forbidden"}

. When I switch between using

start

and

deploy

it will fix the issue for one lambda, but it at times it causes other service to return the same message. The logs for the API gateway don't really show anymore details then returning the 403. My setup is with three stacks. • StorageStack: DynameDB setups here • ApiStack: REST services and use one Lambda Authorizer for all services • WebSocketStack: Websocket rervices Any tips on where I can start to debug this?

Hello. I would like to have your opinion on how you validate (basically which package, if any) requests (or better to say request body) that Lambda receives? (for Node.js)

hey guys! (@Frank @Jay @thdxr) 1 - I’m making a clear separation between my

dependencies

and my

devDependencies

, but on of my dev dependencies, even though it’s not a dependency of any of the dependencies, it’s going to the deployed lambda…is that expected? 2 - is there tree-shaking when deploying the lambdas?

Hi everyone! I just heard about SST today and start following your amazing guides! One quick question on the live lambda development - if my lambda function needs to make a dynamoDB request, then we would have to set up a VPC first, is my understanding correct?

Maybe someone here can help with a CDK problem I am trying to solve. I am currently moving all my code into a mono repo. At the top level of my stack I specify each stack for each environment and then when I deploy I specify the stack. So my toplevel app looks like this:

const app = new <http://cdk.App|cdk.App>();

new BackendStack(app, `BackendStack-dev`, {
...

new FrontendStack(app, `FrontendStack-dev`, {
...

new BackendStack(app, `BackendStack-sandbox`, {
...

new FrontendStack(app, `FrontendStack-sandbox`, {
...

So if I want to deploy to

sandbox

I can do something like this:

cdk deploy --require-approval never BackendStack-sandbox 'BackendStack-sandbox/**' FrontendStack-sandbox

The problem is everytime I synth/deploy CDK synths all stacks before dropping them for final output. The issue is that if I have time consuming operations (such as building assets) in my develop stack, I need to wait for them to complete when deploying to my sandbox stack. A solution I have found is to wrap the above constructs in if statements. This works but now I want to use the stack names I passed to cdk deploy. Does anyone know how I can reference this in code? In the example above I would like to get

BackendStack-sandbox 'BackendStack-sandbox/**' FrontendStack-sandbox

. I thought this would appear in the context but I am not seeing it anywhere.

can u help me to integrate https://www.npmjs.com/package/dynamodb-geo this package with stack to create table with this

Hi everyone, it is my very first post so excuse me if I missed something. I’m interested in calling lambda from another lambda, the 1st one is Node-based and deals with the upload itself + some validation but I would like to invoke another one which is processing the data using Python, I have found this PR about attaching permissions but I got stuck how to proceed further. Thank you

Is it possible to create 2 state machines that reference each other via the startexecution task?

Is it possible to specify log retention for a created stack? All I could find for specifying the log retention for a

SAM

app is this repository in SAR: https://serverlessrepo.aws.amazon.com/applications/arn:aws:serverlessrepo:us-east-1:374852340823:applications~auto-set-log-group-retention So I have two questions. 1. How can I see the current retention for a stack? 2. Can I change it without per stack without specifying every log group for a function one by one?

Looks like the CDK CustomResource changed since I last used it. I things like the CustomResource Provider Framework… that reads like it handles a ton for you. For instance, a lot of examples show to implement the CR functions as

lambda.SingletonFunction

which makes a ton of sense if you are building a construct that could be used a number of times in the same stack…. you would not want it to reimplement the CR function as many times. However, I’d like to be able to use the same asset compression for functions that

sst.Function

gives you. Am I missing something, is there an easy way to have my cake and eat it too?

Hey Guys, The external REST service I am calling takes 35-50sec to respond. I am thinking of replacing API Gateway with an Application Load Balancer. I know it's not serverless and generates a monthly bill, but I can't find a way around 30sec timeout limit at APIG. I am not sure if anyone has tried this yet but I have a couple of questions: • Should I move all routes to ALB or just the ones which are getting timed out? Since I will be paying the ALB bill anyways does it make a difference? • Should I try decoupling my API into 2 parts? It's basically a booking service. Instead of sending a req and waiting for 35-55 sec, Should I submit a job and poll for status from UI like every 10 seconds? ◦ APIG calls lambda which pushes input to SQS and returns a polling ID ◦ SQS calls the consumer lambda, which calls external service, and writes data to DB against polling ID. ◦ Another APIG route takes polling ID and checks if the record is available in DB.

Hello. I want to send “unsuccessful” Lambda invocations by APIGateway to somewhere. I guess I can define “unsuccessful” as it an invocation resulting in

4XX

and and

5XX

status codes. It’s possible that a runtime error thrown, or simply ran out of memory, or timed out. But at the end, it is still the case that APIGateway responds, in this case the statusCode would be some 5XX’s, so I would know that there is something wrong with the Lambda. As I read

AWS::Serverless::Function

resource properties, I see that there are two properties might be relevant, but I’m really not sure how they differ, and which one to be chosen. What is really the difference between

DeadLetterQueue

vs

EventInvokeConfiguration

Hey guys, looking at Cognito User Pools for the first time and I’ve clocked that every time we make a change, even though the Auth construct remains (thanks to default RemovalPolicy), the client id and secret get regenerated. Is there any way to keep those static between updates? Had a quick look through the type def. files and through some of the sst docs but couldn’t find anything about it. I’m gonna go and look at the CDK docs next but thought I’d reach out in the meantime.

Hi all, question on CORS. Why do we have to configure it in both Api construct and lambda responses? Are they for different purpose?

How can we get stack outputs into our CI variables?

Hi everyone, there was a mention of SST supporting python applications, unless I misunderstood that. Is there a guide i can follow to set one up?

Does sst CLI have a

--no-rollback

option like cdk CLI does?

Does anyone happen to know if its possible to do a loop in step functions given a number for the amount of times it should loop rather than a list of items? I know I could use a lambda to transform into an array but if I can do it without that would be nice.

Is there anyway to setup policies via SST/CDK to tear down a stack after a certain amount of time? Or how are others handling making sure ephemeral stacks (feature branches) are not sitting out there forever?

Hello everyone, first of all thank you for working on this amazing piece of software ❤️ I am currently stuck trying to deploy my first stack, a simple hello world. When running 'npx sst deploy' I get stuck at the MFA phase, I put in the code(nothing shows, like a sudo password prompt), I press enter, nothing happens.

Hello everyone! Basic question: after

npx sst remove

I went into AWS and "cleaned up" whatever wasn't removed... apparently not a good idea, because if I now run

npx sst start

I get the following error:

Deploying stacks

dev-1xxxxx-debug-stack: deploying...

[0%] start: Publishing xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx:current

[50%] fail: No bucket named 'cdk-xxxxxxxxxxx-assets-xxxxxxxxx-us-east-1'. Is account xxxxxxxxxxxxxxxxxx bootstrapped?

[50%] start: Publishing xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx:current

[100%] fail: No bucket named 'cdk-xxxxxxxxxxxxxx-assets-xxxxxxxxxxxx-us-east-1'. Is account xxxxxxxxxxxxx bootstrapped?

There should be a simple solution to this, unfortunate wasn't able to find any answers online, appreciate any help!

So, I’ve got a bit of a weird one and it’s probably more CDK related, but I’ve decided to split out my Cognito User Pool (and client) code in to it’s own stack to make some planned work more manageable. In this stack I’m create a config object for my API in another stack. I’ve tried it a couple of different ways and I’m getting a circular reference detected when CDK tries to synth. My set up: • index file creates cognito stack • cognito stack takes the app in constructor (like all stacks do) • cognito stack set up the user pool/client • cognito stack builds a config using some of its resources values • index passes the expose config to my api stack The cognito stack doesn’t reference anything from outside of itself and only uses the

app

var to pass to the sst.Stack

super

. Anyone else come across anything like this before?