I posted this in a dedicated AWS workspace to no avail so posting here too in case anyone has any ideas I want to be able to provision users with a standard set of policies across multiple accounts created using Control Tower Account Factory. I.e. there'd be a user in each AWS account that had the same centralised set of policies applied. It seems like I need to use a Permission Set in SSO that can then be applied to a user when it's created. However, the only options that gives me is to select up to 10 AWS Managed Policies or create a single custom policy. The policies we want to apply are quite detailed as they're to enable a user to deploy a variety of AWS resources for a serverless deployment. It appears that if we go this route that we'd need to paste the content of all of our policies into a single Permission Set policy. I haven't tried yet as this seems a bit weird to me but I'd be surprised if the size of it was allowed. Am I thinking about this wrong or missing something? Any guidance greatly appreciated.

Is there a good resource for EventBridge that goes beyond what is is and why it's great ? Specifically interested in dev experience, how to manage schemas and if there's tooling to help.

DAE: Submit a reply to an AWS support case, get an email few minutes later with “New correspondence added”, but when you go to check it is your own reply 🤦🏽‍♂️….gets me every time!

Looking forward to seeing SST at next year's cdkday 😄

What are the biggest products you guys know of that are using full stack serverless in a meaningful way?

Hello everyone! What are your opinion about Middy (https://github.com/middyjs/middy ) ? Ever tried? I’m using in a few projects and loving it. I even release my second (is just one my first steps in open source - fourth to be exactly) middleware: https://dev.to/aws-builders/redis-exploring-redis-as-serverless-database-to-solve-idempotence-in-apis-2gma

Hi all! Anyone able to give some DynamoDB advice to someone very fresh to the NoSQL world? Maybe there's a good resource you can point me to. I've looked at dynamodbguide and watched some videos but I'm still not grokking it I'm creating a site that has: • User profiles • Users create talent trees (think this) • Users vote on "leaves" within those talent trees My thinking to implement the DB is: • PK: USER#name • SK: ◦ PROFILE#name ◦ TREE#id Attributes for the TREE#id item would be name, createdAt etc, but this leaves me with the following questions: 1. how should I go about storing the tree hierarchy itself (json stored within the TREE#id?) 2. where should the votes be stored? Directly on the tree? 3. down the line i'd like to be able to have users revisit an old tree and see what they voted for... does this change the implementation?

a cool sst github action would be a nice to have: that automatically deploys your github repo for every push or merge.

Any Notion fans out there? We are an early stage startup conducting market research on ways to enhance Notion for those who use it most. All I ask is for 15 minutes to learn your current use cases and future needs. Please DM me and let’s talk!

Hey guys, I have a random question not related to sst or serverless but github Action. in my github Action.. i have a build step(for react app) this step uses environment variables (Action secrets). I dont want to generate build 3 times, for dev staging and prod. so i am thinking to always generate build with placeholders and use sed to replace the environment variables in the build right before deployment. what could be the possibile drawbacks of doing this. below is a command that i came up with (run this inside the <build folder> which is generated after build)

find . -type f -exec grep -Iq . {} \; -exec sed -i 's/Harshit/harshit/g' {} +

did I just break NPM? 😛

how can you break something that's always broken

Cloudflare Workers proxying to Lambda is almost 10x cheaper than API Gateway, and more awesome 😛 Don’t give Bezos all your money

how's everyone doing exception reporting with their serverless setups? Right now I'm using sentry and wrapping all my lambdas but wonder if there's a better way

Related note - what are people doing for handling DLQs? Ideal tool would act like sentry for SQS - alert me sanely, let me inspect failed messages, save them off the queue and let me resubmit pending messages back into the original queue - anything out there like that?

Hey everyone - we’re hiring over here at Fragment 🙂 We’re a ledger API for fintechs, are backed by some amazing investors - Gradient (Google), Box Group and angels from a bunch of fintechs - and our stack is serverless (SLS → SST). More details here and feel free to DM if you have any questions

Just want to give a shout out to this community. My startup, Cloakist, is heavily based around Lambda functions. Using sst makes my life as an indie hacker (team of one) so much easier than it should be. And the support I've got from the people here, especially @Frank, is absolutely nuts. What a cool product, and what a great team! I compare sst/Serverless Stack favourably with another product I use, Supabase - which I also absolutely love, but which doesn't have the completely awesome option of posting for help in a Slack channel and which I therefore find myself way more frustrated with when I'm blocked.

What are people seeing in terms of response times from api gateway + lambda? I have an endpoint that's only doing a simple read from Dynamo and I'm seeing it take 100-200ms. Was really hoping for much faster

Doing some profiling and curious what latencies other people are seeing. If you have a second can you hit this endpoint: https://2270mgssc9.execute-api.us-east-2.amazonaws.com/debug/scrap Let me know how fast it loads + where you're located. I'm seeing 50-60ms from NY

I was writing up a long issue 😢

I really need to break this bad habit of writing angry console.log messages when I’m frustrated 😄 sometimes they accidentally get committed

Hey Folks!, After drawing a complete architecture of my company client upcoming project which majorly uses flutter and firebase. And my bad is that i just came to know that Firebase is not HIPPA Compliant. 😢 Is their any HIPPA Compliant alternative which replace firebase consider as MBaaS (Mobile Backend as a Service) I am also ready to move on any cloud provider but keeping in mind Budget and Time Constraint. Please suggest me .. Thanks

Funny little story We got CircleCI reaching out to us to win us over from Github Actions. But in actuality we're using Seed to build. Here's a snippet

... Nobody I've ever talked to has a 100% success rate on their default branch. Generally, an 80% success rate on a default branch is considered "elite", so that means 20/100 times a build fails. ...

So with SST and Seed, i guess our build system is in the 95+% success rate, therefore elite.

I saw in one of the issues someone mentioned aws-cdk v2 is gonna be a monopackage. Does anyone know more about this?

For those that are one a very heavy event driven architecture, do you have any advice/tips for documentation of how various stacks, lamdba functions, etc. fit into your overall system and what does when? The company I work for is about to hire a handful of engineers as we grow and us two current devs are just thinking through onboarding and long term maintainability and explaining the whole ecosystem to new hires and reduce domain knowledge that just lives in our heads. Any tips and thoughts would be appreciated!

have a great weekend, everyone! 👋 as always, thank you to everyone who answered my never-ending questions 😄 made great progress this week on migrating from multiple Serverless Framework repos to a single SST monorepo

Just for curiosity’s sake, what if I run

npx sst start

for a production application that has requests incoming @ 10/s. I understand that the latency would increase but would my local machine be able to deal with it? FYI - the machine is a Macbook Pro with 2.3GHz 8-core i9 with 16GBs of RAM. Also, if I dare to attempt this what could I do to better handle the incoming load ?

Don’t share this with AWS, those buggers might take it seriously and deprecate cloudformation just to bring something new. 😂

am I way off-base here? this just seems like a huge flaw in CDK. https://github.com/aws/aws-cdk/issues/15929