https://serverless-stack.com/ logo
Join Slack
Powered by
# sst
  • ö
    Hello, am I mistaken that the API url should be 
    <https://blabla.com/dev>
    when you run 
    npx start --stage dev
    ?
    t
    • 2
    • 11
  • t
    Hey random thought, should the StaticSite construct have the ability to spawn a local dev server when running sst start?
    t
    • 2
    • 5
  • j
    Hello all, is there any option to run the API routes from the NextjsSite construct into regular lambdas?
    y
    • 2
    • 1
  • j
    I’d like to use RDS + RDS Proxy which requires to have my lambdas in the same VPC
    t
    • 2
    • 3
  • j
    Maybe your API construct coupled with the StaticSite construct is more what I’m after 🤔
    t
    • 2
    • 4
  • j
    Hey guys, I'm on 
    0.46
    , I was wondering if the option 
    minify
    on sst Function was removed or at a different place. I couldn't it find on 
    Function.d.ts
    t
    • 2
    • 12
  • r
    I'm trying to add Epsagon as a layer as per the instructions here. My code looks like this:
    Copy code
    if (!scope.local) {
      console.log(`Adding epsagon layer from ${process.env.EPSAGON_LAYER_ARN}`);
      const epsagon = LayerVersion.fromLayerVersionArn(this, 'EpsagonLayer', process.env.EPSAGON_LAYER_ARN);
      scope.addDefaultFunctionLayers([epsagon]);
      scope.addDefaultFunctionEnv({
        EPSAGON_TOKEN: process.env.EPSAGON_TOKEN,
        EPSAGON_APP_NAME: process.env.EPSAGON_APP_NAME,
        NODE_OPTIONS: '-r epsagon-frameworks --enable-source-maps',
      });
    }
    It's deployed via seed and I see the console.log inside the if block with the correct layer ARN but no layer is added to any of the functions. What could be going on?
    t
    a
    • 3
    • 46
  • ö
  • m
    Encountered an issue today where live lambda was working but the deployed version failing. The culprit was the graphql module. Setting the NODE_ENV to production fixed the problem. I had to look in the compiled js file from a build to figure this out. Thought I'd share because the way live lambda is run vs. how the code is deployed is not the same and can result in some confusion. 
    Error: Cannot use e "__Schema" from another module or realm.
    k
    f
    • 3
    • 32
  • t
    Yo! Update to my stack failed, when I checked in CF what happened I saw this, should I be worried?
    l
    f
    • 3
    • 3
  • r
    I think previously if there were multiple separate applications running sst start then you'd get asked to specify a new port but just now (on latest version) it bombed out with
    Copy code
    Error: listen EADDRINUSE: address already in use :::12577
    at Server.setupListenHandle [as _listen2] (net.js:1320:16)
    at listenInCluster (net.js:1368:12)
    at Server.listen (net.js:1454:7)
    at Function.listen (/Users/rosscoundon/Documents/GitHub/wheres_my_tech/node_modules/express/lib/application.js:618:24)
    at Server.listen (/Users/rosscoundon/Documents/GitHub/wheres_my_tech/node_modules/@serverless-stack/cli/node_modules/@serverless-stack/core/dist/runtime/server.js:78:18)
    at startRuntimeServer (/Users/rosscoundon/Documents/GitHub/wheres_my_tech/node_modules/@serverless-stack/cli/scripts/start.js:311:10)
    at Object.module.exports [as start] (/Users/rosscoundon/Documents/GitHub/wheres_my_tech/node_modules/@serverless-stack/cli/scripts/start.js:154:9)

Waiting for the debugger to disconnect...
fatal error: all goroutines are asleep - deadlock!
                                                                                                                                        
goroutine 1 [semacquire]:
sync.runtime_Semacquire(0x0)
        runtime/sema.go:56 +0x25
sync.(*WaitGroup).Wait(0xc000180000)
        sync/waitgroup.go:130 +0x71
main.runService(0x1)
        <http://github.com/evanw/esbuild/cmd/esbuild/service.go:138|github.com/evanw/esbuild/cmd/esbuild/service.go:138> +0x4cb
main.main()
        <http://github.com/evanw/esbuild/cmd/esbuild/main.go:200|github.com/evanw/esbuild/cmd/esbuild/main.go:200> +0x1eb

goroutine 34 [chan receive]:
main.runService.func1()
        <http://github.com/evanw/esbuild/cmd/esbuild/service.go:66|github.com/evanw/esbuild/cmd/esbuild/service.go:66> +0x4b
created by main.runService
        <http://github.com/evanw/esbuild/cmd/esbuild/service.go:64|github.com/evanw/esbuild/cmd/esbuild/service.go:64> +0x212

goroutine 35 [chan receive]:
main.(*serviceType).sendRequest(0xc0001be140, {0x13f20a0, 0xc0029c4480})
        <http://github.com/evanw/esbuild/cmd/esbuild/service.go:163|github.com/evanw/esbuild/cmd/esbuild/service.go:163> +0xda
main.runService.func2()
        <http://github.com/evanw/esbuild/cmd/esbuild/service.go:92|github.com/evanw/esbuild/cmd/esbuild/service.go:92> +0x39
created by main.runService
        <http://github.com/evanw/esbuild/cmd/esbuild/service.go:89|github.com/evanw/esbuild/cmd/esbuild/service.go:89> +0x313
    t
    f
    • 3
    • 3
  • s
    @thdxr so, this is very interesting.. https://github.com/Brooooooklyn/swc-node those speeds are insane! maybe this could be worked into SST for unit tests? (there’s a 
    @swc-node/jest
    package)
    t
    a
    +2
    • 5
    • 22
  • ö
    Is it possible to set AWS Profile somewhere in the project? There would be better for people using multiple profiles.
    a
    s
    f
    • 4
    • 31
  • s
    whoa, the 
    .build
    folder can get quite huge over time 😳
    Copy code
    111.4 GiB [###########] /.build
    a
    a
    f
    • 4
    • 8
  • a
    Is there a way to run just the LINTER from SST?
    f
    t
    • 3
    • 9
  • m
    Hey everyone, I have another question. I am using dynamoDB in the backend and I want to support functionality where a user can see how many "total" replies exist for a specific post. dynamoDB doesn't provide 
    count(*)
    functionality. So the only other cost effective option is to trigger kinesis stream on each insert in db and use lambda to aggregate the count. Now here are my 2 questions 1. Is this the best approach? 2. Is there an example i can use which can guide me how to setup kinesis stream on dynamodb with a target lambda?
    a
    a
    +7
    • 10
    • 25
  • d
  • s
    docs request: it would be awesome if the docs had the minimal IAM permissions needed to deploy an SST project
    b
    t
    +2
    • 5
    • 15
  • a
    hi all, is it expected that dynamo tables and cognito user pool wouldn’t be deleted with 
    sst remove
    ?
    s
    f
    • 3
    • 10
  • a
    Somebody shared with me some documentation before about doing an SSH tunnel for MySQL connection of my local lambda… Anybody has it handy? I want to tunnel my SQL over SSH.
    t
    s
    • 3
    • 8
  • a
    Does SST already defines this env-var anywhere?
    Copy code
    [dotenv][DEBUG] "AWS_NODEJS_CONNECTION_REUSE_ENABLED" is already defined in `process.env` and will not be overwritten
    t
    • 2
    • 2
  • a
    Would be possible to add a command to clean all the “.build” folders? Like.. 
    sst clean
    ?
    t
    • 2
    • 4
  • d
    @Frank When building a construct, assuming one that could be added to SST, is there a way to use 
    sst.Function
    where you reference the assets using 
    path
    like you guys do with 
    lambda.Function
    ? Just trying to envision portability of a custom construct.
    t
    f
    • 3
    • 18
  • s
    Anyone tried the Next.js construct against Blitz.js?
    f
    • 2
    • 1
  • d
    @Frank I think there is a small bug with the debugger incorrectly not detecting an infrastructure change. I changed the 
    copyFiles
    from
    hit save and it said no changes detected.
    Copy code
    new sst.Api(scope, 'AppAPI', {
      defaultFunctionProps: {
        bundle: {
          copyFiles: [{ from: '../../../contentful-app-frontend', to: '.' }],
        },
      },
      routes: {
        'ANY /': 'lib/constructs/ContentfulApp/lambda.handler',
        'ANY /{proxy+}': 'lib/constructs/ContentfulApp/lambda.handler',
      },
    })
    t
    • 2
    • 9
  • m
    Hey guys, dumb question. I am new to this javascript/typescript paradigm and one of the difficulty that i am facing is lack of autocomplete/suggestions in visual studio code. I used to use Intellij for java and it used to give hints what public fields/functions/function parameters are available/required etc. Is there anything in vscode that i can install that can give me same intellisense for SST (typescript)
    t
    r
    +2
    • 5
    • 5
  • e
    Hello all, I'm using sequelize (with mysql, and NOT postgres) and getting the following error on deploy (it works ok locally without any bundle props) : 
    Building Lambda _function_ src/list.main
    > node_modules/sequelize/lib/dialects/postgres/hstore.js:3:23: error: Could not resolve "pg-hstore" (_mark_ it _as_
    external
     
    to
     
    _exclude_ it _from_
    the
     
    _bundle_, _or_
    _surround_ it _with_
    _try_/_catch_
    to
     
    handle
     
    the
     
    failure
     
    at
     
    _run_-_time_)
    3 │ const hstore = require('pg-hstore')({ sanitize: _true_ });
    And I've tried putting the module as external_modules as recommended (and also tried node_modules) but with no success in my stack API definition: 
    defaultFunctionProps: {
    bundle: {
    node_modules: ['pg-hstore'],
    // external_modules: ['pg-hstore'],
    // node_modules: ['pg', 'sqlite3', 'tedious', 'pg-hstore']
    }
    },
    Any idea ?
    t
    • 2
    • 17
  • g
    Ok. I gotta say, this looks too good to be true. We are getting our toes in the water with the sample Notes app, but are very pleased thus far. For years, I've been hearing how NoSQL is the way to go and I just cannot get my 56 year old RDB head around that idea. We deal with extraordinarily complex SQL queries and I am on about year 10 of trying to warm up to the idea that such things can be done efficiently with something as simple as key/value pairs. Every time I look into a demo, it's always a Hello World sample of DynamoDB or MongoDB and doesn't address the challenges of working with large and complex datasets with foreign keys, one to many relationships, 4th normal form, etc. But, I digress. The idea of having something like SST abstract away all of the Lambda, CodeDeploy, SQS, S3, CloudFront, DyanmoDB, etc. It's just overwhelming. Realizing that we can move forward on projects without needing to spin up, maintain, and understand EC2 Linux servers... Well, it's difficult to articulate how big a deal I think this is.
    s
    t
    j
    • 4
    • 8
  • a
    Is somebody using AWS Code Artifact? I want to host some private NPM packages but I don’t want to pay to NPM.COM
    t
    j
    • 3
    • 7
  • d
    How do I get SST to build with proper sourcemaps? My stacktrace in seed.run issues are minified.
    t
    t
    • 3
    • 35
1...161718...33Latest