Definitely something I am interested in following especially to see how development directions go, hope the development team is watching and can find this feedback to be useful.
This is just simply evil. Cloud resources created out of nowhere w/o any review and underline IaC. It's something that will give both infra and security headaches. Maybe it makes sense for dev/sandpit environments but those are subject to security reviews as well. I mentioned it from the start when I saw Pulumi in action: yeah it will give out more liberty to the developer, but it will open a door to havoc, just like Jenkins scripted pipelines did long time ago and will bring unnecessary complexity into the equation. Terraform (and up to some point CloudFormation) are tools with a short learning curve thanks to DSL/HCL which give more control on the underlying infrastructure and even overall release process. Division of labor applies in SD too, that's why there are DevOps, Infra, DevSecOps etc. These kinds of tools are not for everyone. In my experience most developers don't know squat or don't care at all about infra and sec, they just want to deliver. Also from my experience, every time a dev gets access to Create* a resource in the cloud, it's either not encrypted or in a public network w/o any security rules and so on.
There is an underlying IaC. It currently uses Pulumi (others are promised) and everything is generated in the local file system. Everything in that output is Open Source. So, you can store it in git, review it, or do anything else you would normally do. On top of that, my understanding is that the ability for ops to create some sort of rules that will govern the output is in progress.
Very interesting! Kind of reminds me of something like springboot on Java where annotation end up doing some behind-the-scenes work. For simple things this may make sense, but there is a part of me that feels this will be unable to fit the needs of a large portion of development. The other thought that occurs to me, is that if this takes off, AWS / GCP may put something out like this on their end with similar purpose which would also be interesting to see.
I think that it all depends on how extensive it becomes. In theory, in-code annotations can be used to generate any type of infrastructure or services. The main question is whether the project will grow to accommodate different needs and situations and, if it does, whether the additional complexity will make it less user-friendly. It's in the early stages so, for now, for the majority of us, it's mostly about waiting to see what'll become.
I can definitely see this as something that can be really helpful in a "dev environment" type manner where you can quickly equip developers to get their code out into the cloud to utilize serverless and Containers as a Service solutions in different cloud providers that are supported at this time. That said, I do wonder how much will be required for a full production lifecycle of applications to be supported by this.
I do think getting to the point of having "a full production lifecycle of applications to be supported by this" is possible in the future. The project is not there (it just started) so we'll have to wait and see whether it does indeed get there.
We're already in production with a few smaller startups we work with - we're slowly (and safely) ramping up towards larger workloads. Our team has extensive expertise in large scale development and operations. We're all excited to see how fast developers can build with Klotho
Thanks for share it! No doubt It is a very innovative idea. For companies it would be difficult to estimate costs relying the resources to be created to a tool that creates what it consider necessary... It would be nice to see the improvements in years ahead
Thanks Javier! We're working towards incorporating/integrating with cost estimation tools that can (roughly) project costs based on different parameters you set
Pretty cool. Great I would say for very simple apps or business new to cloud to get things moving. I would think seasoned enterprises that work with Cloud/AKS/EKS etc regularly - They would prefer the control over their services with git-Ops, etc.. Very cool though - I like the idea of not having to worry about everything.. I wonder though, once things change how will it manage migrations and updates..
The plan, as far as I understood, is to give ops means to make it tailor-made for what a company needs. Nevertheless, that's a plan and not yet a reality.
What do you think about Klotho? Is that a kind of a project in early stages that might be worth following?
Definitely something I am interested in following especially to see how development directions go, hope the development team is watching and can find this feedback to be useful.
@@danielsong284 We're definitely watching 😊
This is just simply evil. Cloud resources created out of nowhere w/o any review and underline IaC. It's something that will give both infra and security headaches. Maybe it makes sense for dev/sandpit environments but those are subject to security reviews as well.
I mentioned it from the start when I saw Pulumi in action: yeah it will give out more liberty to the developer, but it will open a door to havoc, just like Jenkins scripted pipelines did long time ago and will bring unnecessary complexity into the equation.
Terraform (and up to some point CloudFormation) are tools with a short learning curve thanks to DSL/HCL which give more control on the underlying infrastructure and even overall release process. Division of labor applies in SD too, that's why there are DevOps, Infra, DevSecOps etc.
These kinds of tools are not for everyone. In my experience most developers don't know squat or don't care at all about infra and sec, they just want to deliver. Also from my experience, every time a dev gets access to Create* a resource in the cloud, it's either not encrypted or in a public network w/o any security rules and so on.
There is an underlying IaC. It currently uses Pulumi (others are promised) and everything is generated in the local file system. Everything in that output is Open Source. So, you can store it in git, review it, or do anything else you would normally do. On top of that, my understanding is that the ability for ops to create some sort of rules that will govern the output is in progress.
This looks very promising! Thanks for bringing this up!
Very interesting! Kind of reminds me of something like springboot on Java where annotation end up doing some behind-the-scenes work. For simple things this may make sense, but there is a part of me that feels this will be unable to fit the needs of a large portion of development.
The other thought that occurs to me, is that if this takes off, AWS / GCP may put something out like this on their end with similar purpose which would also be interesting to see.
I think that it all depends on how extensive it becomes. In theory, in-code annotations can be used to generate any type of infrastructure or services. The main question is whether the project will grow to accommodate different needs and situations and, if it does, whether the additional complexity will make it less user-friendly. It's in the early stages so, for now, for the majority of us, it's mostly about waiting to see what'll become.
I can definitely see this as something that can be really helpful in a "dev environment" type manner where you can quickly equip developers to get their code out into the cloud to utilize serverless and Containers as a Service solutions in different cloud providers that are supported at this time.
That said, I do wonder how much will be required for a full production lifecycle of applications to be supported by this.
I do think getting to the point of having "a full production lifecycle of applications to be supported by this" is possible in the future. The project is not there (it just started) so we'll have to wait and see whether it does indeed get there.
We're already in production with a few smaller startups we work with - we're slowly (and safely) ramping up towards larger workloads. Our team has extensive expertise in large scale development and operations. We're all excited to see how fast developers can build with Klotho
Thanks for share it! No doubt It is a very innovative idea. For companies it would be difficult to estimate costs relying the resources to be created to a tool that creates what it consider necessary... It would be nice to see the improvements in years ahead
Thanks Javier! We're working towards incorporating/integrating with cost estimation tools that can (roughly) project costs based on different parameters you set
Pretty cool. Great I would say for very simple apps or business new to cloud to get things moving. I would think seasoned enterprises that work with Cloud/AKS/EKS etc regularly - They would prefer the control over their services with git-Ops, etc.. Very cool though - I like the idea of not having to worry about everything.. I wonder though, once things change how will it manage migrations and updates..
"Microservices are the assemply code of the cloud"
by Victor Farcic
No superise, every time u carry your 4k video to supurise me!
That's a good idea but unfortunately, it seems too opinionated and not customisable enough for bigger companies.
The plan, as far as I understood, is to give ops means to make it tailor-made for what a company needs. Nevertheless, that's a plan and not yet a reality.