Docker vs. Firecracker

You are going to be able to find the most resources and examples using Docker whenever you are working with a container orchestration software like Kubernetes. There will always some entropy when you run in a container, a containerized application will never be as purely performant as an app running directly on the OS. However, in most scenarios this loss will be negligible to the time saved in deployment, monitoring, etc.

Incentivized

While deploying workloads in lightweight microVMs presents a couple of perks, Firecracker may not be the best software to handle this. Startup times are slow and scalability is quite limited because of the jailer and virtualization barriers. We have had security breaches on isolated EC2 instances while using Firecracker. It however has a silver lining by improving how serverless functions in container ecosystems are run with their VMMs.

Incentivized

• Kubernetes context switching - very handy
• Integrate well and seamless - No complicated setup and or UX
• Ease of updating images and etc - very important

Incentivized

• It makes it easy to secure virtual machines as they are segmented into microvms with a minimal attack surface area

Incentivized

• It will be nice to see an admin UI for configuration of images
• Container networking can be simplified
• Data storage/volume management can be simplified

Incentivized

• Excludes devices unnecessarily
• Has a slow startup time the basis being to improve security which is quite irrelevant.
• It has limited functionalities on monitoring VMs despite being an open source.

Incentivized

I have been using Docker for more than 3 years and it really simplifies the modern application development and deployment. I like the ability of Docker to improve efficiency, portability and scalability for developers and operations teams. Another reason for giving this rating is because Docker integrates CI/CD pipelines very well

Incentivized

Haven't seen any outages, fatal/unrecoverable errors in my usage so far. Enough said.

Incentivized

Docker Desktop. The CPU high usage is a known issue. Needs fixing. Otherwise, it is great overall. Would not use anything else still.

Incentivized

We need a solution where initially we can use an OS to trigger our pipeline to be used by terraform and then later in ansible. After doing all work it automatically get exited and we can reclaim the space of our VM. So we created a gitlab pipeline and at the initial stage we defined a docker file which will be our base image and we performed all our activities inside that image to build infrastructure using terraform. Integration we have done in our gitlab pipeline and finally we remove the docker image so that the space can be reclaimed immediately.

It is the only tool in our toolset that has not [had] any issues so far. That is really a mark of reliability, and it's a testimony to how well the product is made, and a tool that does its job well is a tool well worth having. It is the base tool that I would say any organisation must have if they do scalable deployment.

Incentivized

• With Docker adoption, were able to reduce 40% of the time taken to complete workspace setup
• Provided the environment packaging / configurability capability which enabled resource sharing and interoperability across teams
• Were able to leverage off the shelf images provided by leading product companies which increased reliability and security exponently

Incentivized

• Low memory overhead on each microVM
• We can run workloads from different customers on the same machine.

Incentivized