Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Why aren't we all serverless yet?

varoa.net

Post reply on HN
Print view View original

31–40 of 137 posts

Re: Why aren't we all serverless yet?

#31

Post by jvanderbot »

danieloj wrote: post #11

I still find the DevEx of serverless terrible compared to the well-established monolith frameworks available to us. The YAML config, IAM permissions, generating requests and responses, it's all so painful to get anything done. Admittedly I speak as a software engineer primarily building CRUD apps, where frameworks have had decades of development. I can see use cases for event-driven applications where serverless may…

I find serverless to be a breeze, with zero sysadmin costs compared to setting up VPS, EC2, doing your own custom monitoring, etc. Each to their own, however.

And gateway+lambda is a near perfect "dumb crud" app, though it is not without a startup cost.

Re: Why aren't we all serverless yet?

#32

Post by fifticon »

The optimistic tone at the start of the article might just be a hallucinatory strawman set up. But as a probably old dog, I fail to see the allure of these technologies(*).

When I read the copy trying to peddle them, to me it sounds quite like someone saying "Heey.. PSST! Wanna borrow 5000$ in cash, I can give it to you right now! Don't worry about 'interest rates', we'll get back to that LATER".

When I build stuff out of 'serverless', I find it rather difficult to figure out what my operation costs are going to be; I usually learn later through perusing the monthly bills.

I think the main two things I have appreciated(?), is

(1) that I can publish/update functions on cloud in 1-5 seconds, whereas the older web services I also use, often take 30-120 SECONDS(not minutes, sorry) to 'flip around' for each publish.

(2) I can publish/deploy relatively small units of code with 'functions'. But again, that is not quite accurate. It's more like 'I need to include less boilerplate' with some code to deploy it.. Because to do anything relevant, I more or less need to publish the same amount of domain/business-logic code as I used to with the older technologies.

Part from that, I mostly see downsides - my 'function/serverless' code becomes very tied-to-vendor. - testing a local dev setup is either impossible or convoluted, so I usually end up doing my dev work directly against cloud instances.

I'm probably just old dog, but I much prefer a dev environment that allows me to work on my own laptop, even if the TCP/IP cable is yanked.

Oh yeah, and spit on you too, YAML :-) They found a curse to match the abomination of "coding in xml languages" of 20 years ago..

Re: Why aren't we all serverless yet?

#33

Post by giancarlostoro »

danieloj wrote: post #11

I still find the DevEx of serverless terrible compared to the well-established monolith frameworks available to us. The YAML config, IAM permissions, generating requests and responses, it's all so painful to get anything done. Admittedly I speak as a software engineer primarily building CRUD apps, where frameworks have had decades of development. I can see use cases for event-driven applications where serverless may…

Serverless can be useful for very specific tasks, such as processing files you upload, things that should happen in the background, but if you already have a simple monolith web app, I don't see why going serverless just to go serverless will help you.

I do see its usefulness, but its not a one size fits all tool.

Re: Why aren't we all serverless yet?

#34

Post by breckenedge »

danieloj wrote: post #11

I still find the DevEx of serverless terrible compared to the well-established monolith frameworks available to us. The YAML config, IAM permissions, generating requests and responses, it's all so painful to get anything done. Admittedly I speak as a software engineer primarily building CRUD apps, where frameworks have had decades of development. I can see use cases for event-driven applications where serverless may…

There is no good reason to build a distributed monolith. You can always think of/design your monolith as a collection of (micro-)services and get the best of both worlds.

I find FaaS best when needing to automate something completely unrelated to what goes in to serving the customer. Stuff like bots to report CWV metrics from DataDog to a Slack channel.

Re: Why aren't we all serverless yet?

#35

Post by qaq »

sitkack wrote: post #6
qaq wrote: post #4

How about cost at scale? Amazon itself shifted Prime Video from serverless to mostly containers and it resulted in huge savings.

Without a link and a breakdown that makes no sense. We switched from blue to square and saw a honey suckle savings. Amazon runs both and serverless is a billing model. Many serverless runtimes consume containers. Serverless, like microservices are a design philosophy.

the pricing is different for the same amount of compute is that news for you? Lambda is priced an order of magnitude higher than fargate which is priced significantly higher than EC2. For small scale workloads your TCO might be lower with higher level abstraction.

Re: Why aren't we all serverless yet?

#36

Post by dolmen »

danieloj wrote: post #11

I still find the DevEx of serverless terrible compared to the well-established monolith frameworks available to us. The YAML config, IAM permissions, generating requests and responses, it's all so painful to get anything done. Admittedly I speak as a software engineer primarily building CRUD apps, where frameworks have had decades of development. I can see use cases for event-driven applications where serverless may…

Or maybe we just lack frameworks that provide the same developer experience but with transparent serverless deployment?

Re: Why aren't we all serverless yet?

#37

Post by nilamo »

baobun wrote: post #10

This misses the main factor, I think: Vendor lock-in. There is no unification of APIs - every provider has their own bespoke abstractions typically requiring heavy integration into further vendor-specific services - moreso if you are to leverage USPs. Testing and reproducing locally is usually a pipe-dream (or take significantly more effort than the production deploy). Migrating to a different cloud usually requires…

There are a few platform abstractions. Quarkus, a Java framework, has Funqy, an extension that abstracts the differences between something like aws Lambda and Knative triggers, and feels quite easy to use.

https://quarkus.io/guides/funqy

Post reply on HN