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

Open Application Model – An open standard for defining cloud native apps

oam.dev

Post reply on HN
Print view View original

11–20 of 47 posts

Re: Open Application Model – An open standard for defining cloud native apps

#11

Post by tmpz22 »

linkdd wrote: post #7
falcolas wrote: post #2

An abstraction (OAM) On the top of an abstraction (K8s Custom Object specification) On the top of an abstraction (Kubernetes, et.al.) On top of another abstraction (Containers) On top of another abstraction (VMs) I wonder if the Dyke-sealing-boy is available for the inevitable leaks? On a slightly more serious note, even though I participated in OAM early in the process, I've never really seen the value in the abstra…

Imagine a developer: > On top of another abstraction (VMs) Who does not need to badly handle infrastructure > On top of another abstraction (Containers) Who does not need to badly handle isolation > On the top of an abstraction (Kubernetes, et.al.) Who does not need to badly handle scalability > On the top of an abstraction (K8s Custom Object specification) Who does not need to badly handle deployment > An abstractio…

> Who just wants to ship their code

That exists, its called a PaaS. Kubernetes was not meant to be a PaaS.

I'm starting to think people create these orgs to forward their own career, products, or companies.

Re: Open Application Model – An open standard for defining cloud native apps

#12

Post by linkdd »

tmpz22 wrote: post #11
linkdd wrote: post #7

Earlier quoted context omitted.

Imagine a developer: > On top of another abstraction (VMs) Who does not need to badly handle infrastructure > On top of another abstraction (Containers) Who does not need to badly handle isolation > On the top of an abstraction (Kubernetes, et.al.) Who does not need to badly handle scalability > On the top of an abstraction (K8s Custom Object specification) Who does not need to badly handle deployment > An abstractio…

> Who just wants to ship their code That exists, its called a PaaS. Kubernetes was not meant to be a PaaS. I'm starting to think people create these orgs to forward their own career, products, or companies.

> its called a PaaS

You mean like Heroku?

Kubernetes allows your ops and developers to collaborate and scale and avoid vendor lock-in.

How does a PaaS (like Heroku?) allows that?

Re: Open Application Model – An open standard for defining cloud native apps

#13

Post by tmpz22 »

linkdd wrote: post #12
tmpz22 wrote: post #11

Earlier quoted context omitted.

> Who just wants to ship their code That exists, its called a PaaS. Kubernetes was not meant to be a PaaS. I'm starting to think people create these orgs to forward their own career, products, or companies.

> its called a PaaS You mean like Heroku? Kubernetes allows your ops and developers to collaborate and scale and avoid vendor lock-in. How does a PaaS (like Heroku?) allows that?

K8s is one of many options to collaborate and scale applications and for many organizations it is not the best option. It also comes with lots of vendor lock-in if you're using a service like GKE, EKS, or AKE, or integrate with any of their loadbalancer or storage solutions!

Oracle and Alibaba building a large abstraction on top of kubernetes, to do something kubernetes was not designed to do (though capable of doing), is not the direction I personally believe distributed systems development should standardize around.

There are tons of self-hosted PaaS options that are not built on top of massive abstractions and provide better functionality to their users. Since you mentioned Heroku i'll mention https://github.com/dokku/dokku.

Re: Open Application Model – An open standard for defining cloud native apps

#15

Post by linkdd »

tmpz22 wrote: post #13
linkdd wrote: post #12

Earlier quoted context omitted.

> its called a PaaS You mean like Heroku? Kubernetes allows your ops and developers to collaborate and scale and avoid vendor lock-in. How does a PaaS (like Heroku?) allows that?

K8s is one of many options to collaborate and scale applications and for many organizations it is not the best option. It also comes with lots of vendor lock-in if you're using a service like GKE, EKS, or AKE, or integrate with any of their loadbalancer or storage solutions! Oracle and Alibaba building a large abstraction on top of kubernetes, to do something kubernetes was not designed to do (though capable of doing…

Yes K8s is one of many options, but one that can translate to many situations. I'd still choose it over anything else when I need more than a `docker run` on a server.

Thanks for the link btw.

Re: Open Application Model – An open standard for defining cloud native apps

#16

Post by tmpz22 »

linkdd wrote: post #15
tmpz22 wrote: post #13

Earlier quoted context omitted.

K8s is one of many options to collaborate and scale applications and for many organizations it is not the best option. It also comes with lots of vendor lock-in if you're using a service like GKE, EKS, or AKE, or integrate with any of their loadbalancer or storage solutions! Oracle and Alibaba building a large abstraction on top of kubernetes, to do something kubernetes was not designed to do (though capable of doing…

Yes K8s is one of many options, but one that can translate to many situations. I'd still choose it over anything else when I need more than a `docker run` on a server. Thanks for the link btw.

No worries and maybe I am being harsh here but this whole thing seems like a couple big providers trying to push their own idioms on the DevOps industry, an industry which is already saturated on dishonest marketing kool-aid.

"Our solution is way better then all those pre-existing solutions!".

No. No it isn't. Stop acting like you're helping me when you're really selling me something.

Re: Open Application Model – An open standard for defining cloud native apps

#18

Post by jimmyed »

falcolas wrote: post #2

An abstraction (OAM) On the top of an abstraction (K8s Custom Object specification) On the top of an abstraction (Kubernetes, et.al.) On top of another abstraction (Containers) On top of another abstraction (VMs) I wonder if the Dyke-sealing-boy is available for the inevitable leaks? On a slightly more serious note, even though I participated in OAM early in the process, I've never really seen the value in the abstra…

Reminds me of SICP.

Re: Open Application Model – An open standard for defining cloud native apps

#19

Post by fengjingchao »

Hi, OAM and KubeVela author here.

We originally created these specification and projects to facilitate delivering applications to Kubernetes without exposing all the nitty-gritty details. There's a lot to do to achieve such goal besides managing application templates, clusters, environments, cloud resources and service bindings, etc. We have seen common patterns rising up within companies of medium and large sizes. That's why we created them to provide common platform for others to build their own application platforms.

Let me know if you have any questions and I will be available here to answer them.

Re: Open Application Model – An open standard for defining cloud native apps

#20

Post by fengjingchao »

linkdd wrote: post #9
juliangamble wrote: post #5

The spec doesn't appear to address the 12-factor app spec which does a similar job in a more general way. https://github.com/oam-dev/spec/search?q=12+factor https://12factor.net/

Doesn't the 12-factor spec just recommend some practices about writing your app, where OAM is about the infrastructure spec?

YESSSS! That's exactly what I want to say as well :D
Post reply on HN