64 points by cloudarchitect 1 year ago flag hide 26 comments
johnsmith 1 year ago next
Fascinating case study! I've been experimenting with serverless for my personal projects and I'm impressed with the results.
theguru 1 year ago next
Really liked the performance metrics mentioned in the case study. Wondering if it's possible to recreate similar results using a different cloud provider?
hdeutsch 1 year ago prev next
The scaling properties of serverless architectures are truly remarkable. Great analysis!
techlearner 1 year ago next
I have been using serverless in some of my solutions too, especially for burst processing and event-driven architectures. It's been a real game changer, truly humble to see your detailed analysis of limits explored in the case study.
hannahw 1 year ago next
That's an excellent question! We did test that, and I will be sharing more on the effects of extremely high loads during the Developer Conference presentation.
jreynolds 1 year ago next
The major drawback I've experienced with serverless is the additional configuration required as compared to running traditional VPS. How did you handle that while conducting the case study?
devopsguru 1 year ago next
Setting up a solid CI/CD pipeline with automated deployments can make managing a serverless infrastructure simpler. It's worth the initial investment!
jreynolds 1 year ago next
Completely agree with you! I'd suggest starting with AWS SAM to build and deploy serverless applications without overly complex steps.
karlplus 1 year ago next
Working with Lambda functions locally has made my onboarding process to the serverless landscape a little more pleasant. Using local API Gateway tools alongside it has accelerated the learning process.
learner 1 year ago next
I love working with AWS SAM with Python. The boatload of libraries and frameworks available Python-side, combined with the simplicity and speed of provisioning makes development a breeze.
transferablelearning 1 year ago next
AWS SAM and Python has been awesome for my backend development. I'm exploring Azure Functions as a secondary option, just to diversify knowledge.
thebeginner 1 year ago next
Let me know if you find any notable pros/cons between AWS SAM and Azure Functions, as I'm going down the same route. Thanks!
transferablelearning 1 year ago next
Will do! I'll make sure to share any findings in this thread as I learn more! Thanks.
thebeginner 1 year ago next
Appreciate your promise to share, @transferablelearning ! I'm actually trying Azure Functions for the first time now and would learn a lot if you have some insights as you dive deeper!
nobodyprograms 1 year ago prev next
I've heard about serverless, but I am yet to explore its limits. Link to the case study here would be helpful!
hannahw 1 year ago next
Sure, here's the link to the case study: <https://example.com/serverless-case-study>
curiousgeorge 1 year ago next
Thanks, @hannahw ! Read the case study, and I must say it's well-written! I am curious to see if running on 100K simulated concurrent users causes any limitations?
curiousgeorge 1 year ago next
Ah, I'll make sure to catch the presentation then. Looking forward to any additional insights you find interesting!
curiousgeorge 1 year ago next
Exciting stuff to explore! Do you have any tips for those who want to dive into serverless, but don't know how/where to start?
hannahw 1 year ago next
Begin exploring with small, non-critical projects to grow familiar with concepts, tools, and best practices. Look into Serverless Framework and AWS SAM for efficient and consistent development!
thebeginner 1 year ago next
Thanks for the helpful tips! I'm looking forward to understanding its limitations and exploring its benefits for my upcoming projects.
curiousjoe 1 year ago next
I'm assuming you've explored both the NodeJS and Python alternatives for AWS Lambda when conducting the case study. Were there any significant performance differences between the two?
hannahw 1 year ago next
During our evaluation process, we found that NodeJS slightly leads in performance, but the gap shrinks as workload grows. Python worked great for our use-case, and its clearer syntax allows for better readability and maintainability.
curiousgeorge 1 year ago next
I'm using Azure Functions by reusing the Python-based code in my automated tests. How did you create your automation pipelines when testing the limits of the serverless platforms?
hannahw 1 year ago next
We used continuous integration pipelines running against git branches, allowing us to test Lambda performance limits with different scenarios, constraints, and architectural choices. Fully automated across 4 cloud regions!
jreynolds 1 year ago next
Terrific answer! I've adopted a similar approach for my personal projects using GitLab CI/CD pipelines on both AWS and Azure. Keeping the pipelines modular and metadata-driven makes the code much easier to maintain and evolve.