I don't understand the knee-jerk opposition people have to serverless architectures. I recently developed a service[1] with the serverless framework and it was the first time I enjoyed developing server-side code since the era of PHP on shared hosts, where you could just upload code and refresh the page. There's something freeing about never having to think about the server process or what happens if the server is power cycled.<p>So congrats on the funding, I hope you can convert some haters :)<p>[1] obligatory plug: <a href="https://notify.run" rel="nofollow">https://notify.run</a>
I ended up using lamda for several things ad hoc and I have to say the experience is great to quickly add functionality to specific niche things you don't necessarily want to run on your API servers because of weight or simply that having a trigger built in makes the whole flow simpler. However, the downside is if you do something stupid like create an infinite trigger accidentally your bill will exponentially increase. Always remember that part. On a $5 digital ocean instance, they will never charge you $3000 a month for accidentally doing something stupid but AWS will forgive you one time at least and I have my one time now. The most hilarious part of this whole thing...and this is really one of the main points, the $5 digital ocean kuejs (nodel.js) server instance that my AWS lamda was smashing the shit out of with millions of requests did not go down the entire time although had some intermittent slow downs of course. $5 goes a long way apparently.
Congratulations to Austen and Serverless Inc. for raising an additional 10m dollars. From what I've seen they are all very nice people over at Serverless.<p>I've been running a small service similar to their new "Serverless Platform" for some time and was approached by them in 2014 to see about joining their team.<p>Ultimately I ended up deciding not to join because I wasn't convinced there was a strong enough engineering presence in their leadership to make a good product. The next couple of years should be interesting to watch if they can actually build a profitable product.
The coming age of people with no understanding of what running code actually means, no idea how hardware/close to hardware systems behave deep down, is going to be fabulous, and full of wasted computing.
Wooo! I'm happy for them! I am using their stuff right now for a project I'm working on (plug: <a href="https://kurz.app/" rel="nofollow">https://kurz.app/</a>) and I really appreciate the ecosystem serverless is cultivating. Simple stuff like bundling up pip requirements or syncing a local folder with a S3 bucket could be done using a script I write, but through serverless I can just install a package and have it hook in as part of each deploy.
We've been using serverless with AWS lambdas for a few months.<p>Testing is hard, the more AWS shit you tie yourself to the harder local testing and development becomes. I picked up a lambda project another developer started and asked them how they were testing and developing locally. Answer: They deployed a new update for every change (!?)<p>Debugging: Look at log files...<p>Also, at some point serverless added some autocomplete crap to my .bashrc file without asking which I will never forgive them for.
Serverless is great, but I am really loving Zappa for Python Flask and Django development with Lambda and API Gateway.<p>Deployed our first production tool with it and it's been working great.
This looks interesting and I wish them good luck. The problem with any developer tools startup is that no matter how great the product is, the willingness to pay is very low (developers think they can replace you with a small script) and extremaly likely that your market gets slowly eaten from the bottom by open source and/or MS/Google/Amazon fold in your service into their cloud platform.
I usually do TDD serverless apps by debugging unit tests with jest. Is it a bad practice? Anyone can easily mock events
this way, It does not matter whether its a SNS event or an HTTP event. Overall, I have had a great experience with serverless!
Super happy for them. It's clear that lots of companies and people are interested in serverless, as the benefits are clear. this could be how the "microservice" actually manifests itself in a few years.