Hi! Not sure if this is the place to ask such a question but… is there any plans to split the current minimal installation into separate packages? it currently requires tons of packages such as apollo, eslint… even if you only use sst.Function, for example

Hi guys, is there a way to add

replicationRegions: ['us-east-1', 'us-east-2', 'us-west-2'],

in the dynamoDb table construct? seems like replicationRegions is not in the TableProps type.

Played around with the idea of Lazy value, it seems to help with the case where you have an

Api

and a

StaticSite

, where the frontend needs to know the API endpoint, and the API needs to know the frontend URL. Sample code here https://gist.github.com/fwang/db1e5697913c5533f8b95a4f04464870

Does anyone have any tips for troubleshooting a static site that says it deployed fine but when navigating to the sites URL I get: ``````

Hey guys, I’m using Table for DynamoDB and want to setup TTL of 30 days. How to do this with SST/CDK?

/**
     * The name of TTL attribute.
     *
     * @default - TTL is disabled
     * @stability stable
     */
    readonly timeToLiveAttribute?: string;

What should be the value?

Other question guys… I’m using

StaticSite

to deploy my ReactJS site (not using

ReactSite

because my app doesn’t use

create-react-app

framework. So the problem is I’m having some CORS issues downloading fonts from CFront… is there a way to customize CORS for specific file types? Maybe my CORS policy is dictated by S3?

Hi! How do I secure my websocket api created by the debug stack for live lambda development? My actual api gateway endpoints are secured by lambda authorizers, but the websocket api from the debug stack does not have authorizers. I realize it is only a risk when I run sst start locally and the debug session is running. But still if someone somehow knows my websocket endpoint, and the person somehow knows the required values to construct the payload structure as shown here in SST, the person can connect to the websocket api directly to invoke the lambda from my local machine and bypass the authorizer on the api gateway. I tried and it worked. Are there anything in place or any suggestion to secure the websocket api for live lambda development? I know it's a slim risk but I am asking because this is flagged by infosec. If there is no resolution we may have to stop using sst.

I noticed that during the time a

StaticSite

is being deployed the app will not work, I am assuming that some files get changed and some still are the old files and because of the files having the version on the file name (we use vite for building our UI) there is a mismatch and the UI does not load. Has anyone else seen this issue? Any way to prevent this?

hi, I'm currently using v0.60.2 and when I'm attaching permissions to different resources no matter if I remove or add the cli does not detect changes and redeploy event after a ctrl+c and restart, any thoughts on this?

v0.60.6 is also not picking up changes with the .attachePermissions() function

Hi, we discovered some problems with the current versions of sst (I just upgraded from 0.53 to 0.60) and yarn 3: 1. Every time a lambda is invoked in debug mode (

sst start

) for the first time the console is printing strange messages

[Module: null prototype] {
  default: { handler: [Getter] },
  handler: [AsyncFunction: handler]
} handler

…

[Module: null prototype] {
  ALGORITHM_IDENTIFIER: undefined,
  ALGORITHM_IDENTIFIER_V4A: undefined,
  ALGORITHM_QUERY_PARAM: undefined,
  ALWAYS_UNSIGNABLE_HEADERS: undefined,
  AMZ_DATE_HEADER: undefined,
  AMZ_DATE_QUERY_PARAM: undefined,
  AUTH_HEADER: undefined,
  AWS_DEFAULT_REGION_ENV: undefined,
  AWS_EXECUTION_ENV: undefined,
  AWS_REGION_ENV: undefined,
  AccountInfo: undefined,
  AdaptiveRetryStrategy: undefined,
  ArchivalSummary: undefined,
  AssumeRoleCommand: undefined,
  AssumeRoleRequest: undefined,
  AssumeRoleResponse: undefined,
  AssumeRoleWithSAMLCommand: undefined,
  AssumeRoleWithSAMLRequest: undefined,
  AssumeRoleWithSAMLResponse: undefined,
  AssumeRoleWithWebIdentityCommand: undefined,
  AssumeRoleWithWebIdentityRequest: undefined,
  AssumeRoleWithWebIdentityResponse: undefined,
  AssumedRoleUser: undefined,
  AttributeDefinition: undefined,
...

2.

sst update

does not work. Nothing gets updated 3. Seed sometimes complains about the yarn.lock file being updated during a yarn install, which is not the case. This seems to be a caching issue, since a complete redeploy fixes the problems. This only happens when I promote a from one stage to another. We downgraded to yarn 1 again for now. Edit: Only problem 1 is related to the sst version. It’s not happening in 0.53. Problem 2 and 3 have probably always been there under yarn 3.

v0.60.6 is mixing versions of aws-cdk-lib, which breaks everything

% npm ls aws-cdk-lib    
casa@1.0.0 /Users/adam/dev/casa
└─┬ @pkg/cloud@1.0.0 -> ./cloud
  ├─┬ @serverless-stack/cli@0.60.6
  │ ├─┬ @aws-cdk/aws-apigatewayv2-alpha@2.7.0-alpha.0
  │ │ └── aws-cdk-lib@2.9.0
  │ ├─┬ @serverless-stack/core@0.60.6
  │ │ └── aws-cdk-lib@2.7.0
  │ └── aws-cdk-lib@2.7.0
  └─┬ @serverless-stack/resources@0.60.6
    ├─┬ @aws-cdk/aws-apigatewayv2-authorizers-alpha@2.7.0-alpha.0
    │ └── aws-cdk-lib@2.9.0 deduped
    ├─┬ @aws-cdk/aws-apigatewayv2-integrations-alpha@2.7.0-alpha.0
    │ └── aws-cdk-lib@2.9.0 deduped
    ├─┬ @aws-cdk/aws-appsync-alpha@2.7.0-alpha.0
    │ └── aws-cdk-lib@2.9.0 deduped
    └── aws-cdk-lib@2.7.0

dumb question but is there a reason for not using peer dependencies for aws-cdk-lib? is it because you can’t guarantee it will still work for newer versions of the package?

Hey @Frank, Do you have an update on this open issue (speeding up CloudFront update/invalidation)? We're struggling with fairly slow deployments on Seed (~20mins) and about 50% of that time is spent on the 'Deploy' step. Of this, it seems about 3.5 minutes is spent on

AWS::CloudFront::Distribution

. Anything that would help speed up deployments would be hugely helpful for us.

Hi, so I was looking at the SEED output and was wondering how I could break up my services into many services but still have dependencies to some common stack?

INFO: Checking for files changes in the ##### service...
0.07 s
Skipping the check since there is only one service in your app.

Hi everyone. I am trying to run the same SST app on two machines: MAC and UBUNTU. On MAC everything works ok but on ubuntu not all artifacts are created + I am taking the error that a handler is not exported or undefined. Has anyone faced this issue? stage: dev apig: checked. All routes are deployed succesfully

Is there any way to change the handler path for a function post-build. My use case is configuring the New Relic Lambda Extension. This extension expects the function handler to be a method defined in the layer, called

newrelic-lambda-wrapper.handler

. But I cannot set this in my stack code (like in this SAM example), as SST will try to do this and fail:

Building Lambda function newrelic-lambda-wrapper.handler

Error: Cannot find a handler file for "newrelic-lambda-wrapper.handler".

Hello there…. Is there a way to configure the

.env

files to have some level of hierarchy? Like

.env

should be applied always, then on top of it apply

.<http://env.dev|env.dev>

if the STAGE=dev …. and so on… I need to do this due I have some env-vars which are repeated across environments/stages.

Other question… I’m using StaticSite to build a ReactJS app with Webpack (no CRA used). But I can’t get it to receive my environment variables. I ended up doing this:

plugins: [
    new Dotenv({
      path: path.join(process.cwd(), `/sst/.env.${process.env.SST_STAGE}`),
    }),
  ],

But this creates other problems…

[Feature Request - StaticSite]: Can we make deployment prefixes “optional”? https://github.com/serverless-stack/serverless-stack/issues/1323 In-place Deploys: (3mins - avg build time) In the past we have never really been concerned about true blue-green deploys and so our production deploys of our static sites would just use the aws cli to sync and remove orphans with each deploy. sst.StaticSite Deploys: (16mins - avg build time) Contrasting this to us now using 

sst.StaticSite

 and its use of 

deployid

 as a bucket prefix. This now means that the CloudFront distro needs to be updated (origin path) so that adds 6-15mins of replication/deploy time. Optional prefixes and us With optional prefixes, I would like to write a CDK construct that handles Cloudfront distro cutovers and true blue-green deploys. We would only do that for production, while dev deploys stay spicy fast. Even without this construct… one could at lease make the deploy prefixes optional for Prod only, again, keeping dev deploy fast.

Somebody using Epsagon + Middy?

S3 bucket having issues for StaticSite - anybody else seeing this issue?

Another suggestion for a new context, ViteSSRSite. This would expect an app using one the vite SSR libraries (https://github.com/vitejs/awesome-vite#libraries) to spit out a single lambda file for the app. With this you could skip the s3 bucket all together and go with just an APIGateway or maybe this could get really fancy and just use cloudfront and run the Lambda @ Edge. Likely could just wrap up what is being done here https://github.com/jamesladd/aws-vite-ssr. An interesting side effect here, since it is a lambda, is it may be able to work via live dev like other lambdas do...

Do sst nextjs deploys support v12 features (specifically middleware)?

After updating to

SST: 0.61.0 CDK: 2.7.0

I'm now getting

✘ [ERROR] Could not resolve "@aws-cdk/aws-lambda"

    stacks/index.ts:2:24:
      2 │ import * as lambda from "@aws-cdk/aws-lambda"

[SST Console - Feature Request]: 🙏 • Collapse by default (option) - Would love to be able to see everything collapsed by default (or whatever the user decides) so we can see many events come thru and only care about the FAILED ones. • *Collapsable Console Logs*… it seems only the request and response are collapsible but our console logs are expand only. Would love if these followed the same as the point above. • Clear all logs - Sometimes you just want to clear everything away to only see the new stuff.

guys guys.. I found a secret backdoor in SST 😮 (half-kidding) after you run

sst start

, press spacebar and then backspace, and you get a mysterious

>

prompt. doesn’t do anything though actually, @Jay, this is why it takes two ^C’s to kill sst now. once I hit ^C the first time, the weird spacebar/backspace thing doesn’t work anymore. and then the second ^C quits sst. so there’s some kind of prompt process running in addition to

sst start

, maybe?

Hey guys! A simple question. I am trying to add one more account to my aws config/creds and deploy the app to this account but when I hit

npx sst start

I am getting

CredentialsError: Missing credentials in config, if using AWS_CONFIG_FILE, set AWS_SDK_LOAD_CONFIG=1

I have updated

app/.env

file so it looks

AWS_SDK_LOAD_CONFIG=1
AWS_PROFILE=profile npx sst start

but it doesn't work. Am I missing something?

Help, Everything was going so well today and then this:

Checking deploy status...
mike-local-chartflow-dynamodb-stack | UPDATE_IN_PROGRESS | AWS::CloudFormation::Stack | mike-local-chartflow-dynamodb-stack | User Initiated
mike-local-chartflow-dynamodb-stack | UPDATE_ROLLBACK_IN_PROGRESS | AWS::CloudFormation::Stack | mike-local-chartflow-dynamodb-stack | Export mike-local-chartflow-dynamodb-stack:ExportsOutputRefLookupChartflowDataTableB04292358F7B7CAF cannot be deleted as it is in use by mike-local-chartflow-api-stack
mike-local-chartflow-dynamodb-stack | UPDATE_ROLLBACK_COMPLETE_CLEANUP_IN_PROGRESS | AWS::CloudFormation::Stack | mike-local-chartflow-dynamodb-stack
mike-local-chartflow-dynamodb-stack | UPDATE_ROLLBACK_COMPLETE | AWS::CloudFormation::Stack | mike-local-chartflow-dynamodb-stack

 :x: mike-local-chartflow-dynamodb-stack failed: Export mike-local-chartflow-dynamodb-stack:ExportsOutputRefLookupChartflowDataTableB04292358F7B7CAF cannot be deleted as it is in use by mike-local-chartflow-api-stack

I had been working on fixing up my deploys and getting them to work on actual aws hardware and had ignored my local environment. When I went to redploy my local environment this evening, I got this message. I have no Idea what to do to fix. Everything is working great on AWS, now I just have to figure out how I broke my local environment

RE: Lambda ESM Note that I am not looking for help, just sharing Upon conversion to ESM, every graphql resolver fails with the following error:

{"errorType":"Error","errorMessage":"Dynamic require of \"stream\" is not supported","stack":["Error: Dynamic require of \"stream\" is not supported"," at file:///var/task/src/api/graphql/index.js:1:464"

I am not sure this is even the fault of the SST implementation, as it happens in AWS itself, but worth knowing perhaps. I am not doing anything with

stream

directly myself, so it must be inside apollo or another package.