While mowing the yard last weekend, something came to mind about
Fedora on clouds. We
spend a lot of time thinking about getting Fedora onto more clouds and making it more
interoperable with cloud APIs. Increasing Fedora's usage by making it available in
more clouds seems like a worthwhile goal.
This is a great path for your thoughts and I am so grateful that you _do_ think in this
direction when you are really relaxed into a routine.
However, I wonder why people choose to use Fedora versus an
alternative when they deploy
in public clouds. These questions came to mind:
* Is there something lacking in the Fedora experience?
* Is Fedora more difficult to use or does it have limitations that frustrate users?
* Are we missing docs and blog posts that help users deploy their favorite applications
on
Fedora?
I'll admit that I bounce between regular Fedora and Fedora CoreOS in my own
deployments. A lot of that depends on what I plan to run there. If I'm running
containerized applications, CoreOS gets me up and running quickly. If I need to do some
development or run something a bit more complex, I usually reach for Fedora Cloud.
I think that's great. They are both incredibly beneficial, but one doesn't answer
all of my problems. Like you, I feel incredibly comfortable using both FCOS and Fedora
Cloud Editions.
I would love to hear thoughts from others on this topic about how we
can improve the
end-user experience for Fedora in public clouds of all sizes.
There are a lot of variations in workloads, but one of the things that I run into for user
workloads is an ability to predict how they are going to use the tools they have been
using previously in the context of their old process and still deliver big engineering
goals.
One of the things that makes me excited is building supported models for specific tasks.
It's how I built out support for configurations including Nice-DCV on AWS which
ultimately became the proof of concept that cemented process for Amazon EC2 Accelerated
Instances on RHEL Workstation. CoreOS wasn't going to make it through the commercial
partner application security team review for the organization that I needed for a hand off
to the engineering team. It was the shortest path to complete the concept work and helps
me to build out the next generation support efforts.
Another thing that I enjoy working with the cloud images for is the consistency with the
Workstation experience from the perspective of the filesystem. Bringing btrfs onto the
cloud image could not happen without strong community control, but it delivers a lot of
flexibility in the configurations and a significantly different experience to the image in
terms of experimental configurations. It allows me to build, store, and restore in ways
that I could not do with a single experimental instance before. That is super helpful
building out new solutions too.
Also, adding packages into the cloud-sig, so that we are working on an ecosystem for the
support model excites me. I don't care if they are beneficial for only CoreOS or
everywhere. I like the idea that we are building towares a shared core of utilities that
supports management of cloud resources and related configurations. Being able to support
the ecosystem is most important for me as a member of the team.
I think that everyone should be focused on application modernization, but not every
problem is going to call for a new solution. When I reach for the tool to get what I want,
I stick to the old adage I learned from my mentor Wayne Walker a long time ago, which is
this: "The right tool for the job is the tool that gets the job done while I still
have a job!" Sometimes, that's means I pull from what I know will work and that
can be different in different situations or it can be, like I mentioned above, to satisfy
external requirements that are not in my control while I still handle the task at hand.