Latest Twitter Threads by @theburningmonk on Thread Reader App

Nov 14, 2024 • 5 tweets • 2 min read

One of the most misunderstood aspects of Lambda is how throttling applies to async invocations.

Or rather, how it doesn't!

🧵 With Lambda's Invoke API, you can choose invocationType as "RequestResponse" (ie, sync) or "Event" (i.e. async).

With sync invocations, throttling limits are checked to make sure you stay within:
* regional concurrency limit
* the function's reserved concurrencydocs.aws.amazon.com/lambda/latest/…

May 7, 2024 • 11 tweets • 2 min read

SNS & EventBridge share many capabilities and can be used interchangeably in many situations.

So, how do you know when to use which? 🧵

#aws #serverless

Instead of purely judging by the difference in capability (which is an important consideration!) e.g., SNS has FIFO.

It'd be helpful to also consider their respective design choices and let them inform your choice.

Oct 17, 2023 • 16 tweets • 4 min read

I have written over 800 technical articles in the last 14 years and they have been read millions of times.

I see a lot of folks making the same mistakes I did early on in my journey.

So here are some principles to help you get better at writing.

🧵 1. 𝗦𝗲𝗹𝗹 𝘁𝗵𝗲 𝗽𝗿𝗼𝗯𝗹𝗲𝗺, 𝗻𝗼𝘁 𝘁𝗵𝗲 𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻

Like that scene in The Wolf of Wall Street where Di Caprio asked Jon Bernthal to sell him a pen.

First, create the demand, then supply the solution.

Nov 22, 2022 • 6 tweets • 1 min read

This is big!

Payload-based filtering was one of the key reasons to choose EventBridge over SNS. This makes SNS a much more viable option in Event-Driven Architecture.

aws.amazon.com/blogs/compute/… Off the top of my head, this would:

1. Allow you to use one SNS topic for many types of messages, previously one tends to have many SNS topics, one for each type of message.

Oct 24, 2022 • 15 tweets • 3 min read

This is a very interesting thread and the surrounding discussions. I've had similar discussions in the past, but I don't want to make too many assumptions here because I wasn't part of the conversation with the client.

But here are my thoughts.

https://twitter.com/BorisTane/status/1583507431564333058

First of all, serverless has been successfully adopted at a much bigger scale in other companies. LEGO for example have 26 squads working mostly with serverless, and PostNL has been all in on serverless since 2018.

So technology is probably not the problem.

Aug 11, 2022 • 12 tweets • 13 min read

Serverless is great, but it's not a free lunch.

The "serviceful" mindset means you'll need to learn and use many services in your architecture. Having the right tool to help you along can make a world of difference. And here are 7 of my favourites. Serverless IDE is a VSCode extension that will save you countless hours when you work with CloudFormation, AWS SAM or the Serverless Framework.

Auto-completion, schema validation, CF docs on hover, and many more.

marketplace.visualstudio.com/items?itemName…

Aug 8, 2022 • 11 tweets • 5 min read

Don't wait for users to complain to find out there's a problem.

Here are the alerts you should have around your #serverless application.

It's not an exhaustive list, but a good starting point that's broadly applicable to everyone. Most of the regional metrics for Lambda aren't useful for alerting you of an emerging issue, EXCEPT the regional concurrent execution count.

Set an alert threshold to ~70% of your current regional limit so you know to ask for a limit raise before it's too late.

Aug 4, 2022 • 15 tweets • 4 min read

I have written ~800 technical articles/whitepapers in the last 10 years and they have been viewed millions of times.

I see a lot of folks making the same mistakes I did early on in my journey.

So here are some principles to help you get better at writing.

🧵 1. 𝗦𝗲𝗹𝗹 𝘁𝗵𝗲 𝗽𝗿𝗼𝗯𝗹𝗲𝗺, 𝗻𝗼𝘁 𝘁𝗵𝗲 𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻

Like that scene in wolf of wall street where Di Caprio asked Jon Bernthal to sell him a pen.

Create the demand, then supply the solution.

Aug 1, 2022 • 17 tweets • 15 min read

Here are 10 performance optimization tips I learned after rebuilding two of my landing pages from scratch.

If you're a backend-focused dev like me, these might be new to you too.

They gave my sites a great perf improvement, especially on mobile.

🧵

1. If you have a logo image above the fold, then tell the browser to prioritize downloading it by adding it as a preload in the <head> section of the index.html

<link rel="preload" as="image" href="... />

Jul 31, 2022 • 5 tweets • 2 min read

I love step functions, but the 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱 workflow's pricing model (based on no. of state transitions) can get hairy at high throughput and/or for large state machines with lots of states.

𝗘𝘅𝗽𝗿𝗲𝘀𝘀 workflows solve the cost issue but you lose a lot in the process. U can't visualize executions, no exactly-once execution, and limited to 5 mins.

A nice pattern is to combine 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱 and 𝗘𝘅𝗽𝗿𝗲𝘀𝘀 workflows and get the best from both by nesting 𝗘𝘅𝗽𝗿𝗲𝘀𝘀 workflow inside a 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱 workflow.

Jul 18, 2022 • 17 tweets • 9 min read

Are you spending too much on Lambda?

Here are some ways to save money on your Lambda bill.

🧵

#serverless #awslambda Lambda is usually very cost-effective, you only pay when they run (with ms billing) and they scale to zero.

But a combination of misconfiguration and a high throughput use case can give you a nasty surprise when you get your AWS bill, esp if you don't have billing alerts set up.

Jun 17, 2022 • 13 tweets • 4 min read

Great question from my current cohort of students, paraphrased:

"Should you always use Step Functions to chain together a few Lambda functions? Are there patterns to simplify this? How about using SQS between the functions?"

Here are my thoughts 🧵 Firstly, on the broader topic of orchestration vs choreography, I've written my thoughts before. TL;DR is that I prefer orchestration for intra-service workflows, and use events for inter-service communication.

theburningmonk.com/2020/08/choreo…

May 5, 2022 • 8 tweets • 3 min read

As the Lambda service becomes more mature and fully featured, there's also more confusion around when to use these new features. Function URL came up in a conversation today, so let's talk about that!

🧵 Let me start by saying that I'm quite excited by its release and I think it's great that it's now an option.

But I also think it shouldn't be the default for most of the people who are using Lambda today.

lumigo.io/blog/aws-lambd…

May 3, 2022 • 19 tweets • 9 min read

The no. 1 question I get about #serverless is around testing - how should I test these cloud-hosted functions? Should I use local simulators? How do I run these in my CI/CD pipeline?

Here are my thoughts on this 🧵 There's value in testing YOUR code locally, but don't bother with simulating AWS locally, too much effort to set up and too brittle to maintain. Seen many teams spend weeks trying to get localstack running and then waste even more time whenever it breaks in mysterious ways 😠

Dec 17, 2020 • 17 tweets • 7 min read

If you want to learn about the internal details of Lambda, then check out @MarcJBrooker's session "Deep dive into AWS Lambda security: Function isolation"

virtual.awsevents.com/media/1_gu6tp8…

Here's my play-by-play

🧵 haha, a typical EC2 bare metal instance can run A LOT of Lambda workers! Which means AWS needs to figure out how to do it securely and performantly.

Dec 6, 2020 • 8 tweets • 5 min read

Stitching together my previous tweets on @dyanacek's session (BLD205) on "Monitoring production services at Amazon" so it's easier to share

#aws #reinvent2020

virtual.awsevents.com/media/1_4nwtxc… "For amazon.com we found the "above the fold" latency is what customers are the most sensitive to"

Dec 5, 2020 • 30 tweets • 12 min read

If you missed @clare_liguori's Continuous Delivery session this week (like I did) then good news, it's available on-demand now 🎊

virtual.awsevents.com/media/1_ua3d99…

And here's my play-by-play for the session

🧵... This is a typical CD pipeline in AWS.

This is far more complex than the most complex CD pipeline I have ever had! Just cos it's complex, doesn't mean it's over-engineered though. Given the blast radius, I'm glad they do releases carefully and safely.

Dec 5, 2020 • 11 tweets • 3 min read

I've gotten a few questions about Aurora Serverless v2 preview, so here's what I've learnt so far. Please feel free to chime in if I've missed anything important or got any of the facts wrong.

Alright, here goes the 🧵... Q: does it replace the existing Aurora Serverless offering?
A: no, it lives side-by-side with the existing Aurora Serverless, which will still be available to you as "v1".

Dec 3, 2020 • 13 tweets • 6 min read

Great overview of permission management in AWS by @bjohnso5y (SEC308)

Lots of tools to secure your AWS environment (maybe that's why it's so hard to get right, lots of things to consider) but I love how it starts with "separate workloads using multiple accounts"

SCP for org-wide restrictions (e.g. Deny ec2:* 😉).

IAM perm boundary to stop ppl from creating permissions that exceed their own.

block S3 public access

These are the things that deny access to things (hence guardrails)

Use IAM principal and resource policies to grant perms

Dec 3, 2020 • 20 tweets • 8 min read

Great session by @MarcJBrooker earlier on building technology standards at Amazon scale, and some interesting tidbits about the secret sauce behind Lambda and how they make technology choices - e.g. in whether to use Rust for the stateful load balancer v2 for Lambda.

🧵

Nice shout out to some of the benefits of Rust - no GC (good for p99+ percentile latency), memory safety with its ownership system theburningmonk.com/2015/05/rust-m… great support for multi-threading (which still works with the ownership system)

Dec 2, 2020 • 44 tweets • 14 min read

This is part 2 of my #aws #reinvent hot takes on the big #serverless related announcements.

Part 1 is here for anyone who missed it:

https://twitter.com/theburningmonk/status/1333350172181622789

Same deal as before, if this gets 168 retweets then I'll do another batch 👍

Alright, here comes the mega 🧵... 1. Lambda now bills you by the ms as opposed to 100 ms. So if your function runs for 42ms you will be billed for 42ms, not 100ms.

This instantly makes everyone's lambda bills cheaper without having to lift a finger. It's the best kind of optimization 😎

aws.amazon.com/about-aws/what…

Share this page!

Enter URL or ID to Unroll