So performance any different except for it’s a huge difference depending on the language, whether it is in VPC forcing the need to create ENIs, etc.
I’m a multi certified AWS true believer and I love lambda for everything else besides APIs, but let’s not ignore the large performance penalty of cold starts and lack of connection pooling.
You really don’t save money over just using a bunch of micro instances behind a Load Balancer and autoscaling. Elastic Beanstalk makes managing it dead simple.
Back then developers were dismayed about the unpredictable cost of App Engine. The same problem remains today with Lambda and others.
Is it just a solution whose time has come now, ten years later? Or is there something else to Lambda that I’m missing?
Also original GAE was hampered by having a product owner that didn't care about it (at the time I was working on the largest GAE app inside Google). We (as internal customers) ending up writing chunks of the AE SDK to be able to have our apps run reliably.
I spent more than that on a single train in to the city today.
Or you can just run from a residential connection for effectively free if your service is _that_ small.
#2 - "Serverless" paradigms change often enough that it's essentially another form of patching.
It's also obviously not true, any system you write code on needs to be patched because it has auth to this system.
#3 seems to be buzzwords with no substance?
As far as I can tell, the only "problem" that "serverless" (i.e. Our Servers) solves is Amazon not monetizing everything ever.
While you're busy figuring out what monitoring solution to use on your VPS to monitor the running all for the sake of a little Python script, I've already set up a lambda function and maybe set a cloudwatch alarm to look after it and can move on to the next thing and leave it running indefinitely. Depending on the usecase I might not even pay a penny during that time, either.
I swear 90% of the hate for serverless purely stems from irritation at the name.
I have checked the definition of the word "editorialise: To present an opinion in the guise of an objective report." - I did not express any opinion in the title - I simply added context for those who don't know about the so-called "contaminated blood scandal" in the UK. Namely the three-letter detail I added was "HIV" i.e. that the UK's government-run National Health Service infected patients as young as 8 years old with HIV.
This change was not gratuitous as it added essential context to the story.
I have checked the Hacker News guidelines and since the article dispels myths people have about health systems, I felt it was pertinent.
If you are spending that much time managing infrastructure on AWS, you’re doing it wrong. ElasticBeanstalk makes it dead simple.
While you're busy figuring out what monitoring solution to use on your VPS to monitor the running all for the sake of a little Python script, I've already set up a lambda function and maybe set a cloudwatch alarm to look after it and can move on to the next thing and leave it running indefinitely. Depending on the usecase I might not even pay a penny during that time, either.
And you can do the same with AWS and a bunch of micro instances. But if it’s that low volume, I would just host it myself.
I swear 90% of the hate for serverless purely stems from irritation at the name.
I love serverless for everything but APIs. But 90% of my hate for serverless is with APIs and the cold starts.
Been there, done that. Buy a raspberry pi, install raspbian, install nginx, node, whatever the stack is gonna be, configure port forwarding on a router, spend hours to days getting it working. Have tons of downtime--oops you don't have a static ip, oops your sd card got corrupted, hey lets update this to the latest version oh no everything is broken...and on and on and on.
Or, figure out how to use S3 buckets and lambda functions once, and from then on watch everything just work and cost virtually nothing. For small (I'm the only user) types of projects, I'll stick to serverless for now on, so I can be a programmer and not a sysadmin.
This is very true. I’m an AWS true believer and I use most of the most popular features all of the time for work but for a low volume side project, I would host it on my own computer. I have 900Mbps upload bandwidth. If I were going to use AWS, I would use a cheap LoghtSail instance.
Different use cases warrant different solutions.
Not gonna be possible soon, now that ISPs are starting to hand out private IPs to customers. That's going to keep happening until the world moves to IPv6.
There obviously are servers, "serverless" just implies you don't own them.
There's a few problems that using serverless solves:
1. It's great for low volume sites or services. You don't need to rent hardware that runs the whole month if you only get pinged a few times per day.
2. It reduces operational burden. A whole class of problem goes away - patching operating systems and application servers.
3. It provides an abstraction layer for processing events. Events could be HTTP requests to a site, data coming in from an IoT device, that a new file was uploaded to S3, etc.
Performance isn't that different from using EC2 (or the non-AWS equivalents), other than cold start times, and choice of language platform can help mitigate that.
In terms of complexity, for the project work I've done the complexity isn't different than running something in any other application server.
Cloud providers other than AWS have a serverless offering, and there are frameworks out there for abstracting the differences, so tooling support doesn't have to come with lock-in. In addition, a well-architected serverless application has a pretty small layer interfacing with the serverless framework, so the code lock in is negligible.