#codefresh (2020-06)

I cross posted this to the community, I’ll update here if I get an answer…but just incase I forget https://community.codefresh.io/t/sdk-pipeline-run-documentation/380

That’s more of a question of “To Kubernetes, or not to Kubernetes”

EKS is Kubernetes, ECS is not

yes, I know

If you already use Kubernetes, then EKS. If you don’t, but want to start, EKS. if you don’t, and don’t want to start, then ECS

looks like unless you’re on the enterprise plan you can only have 1 concurrent runner running

That’s a shame to find out. I was hoping that now that they made this available to everyone that it would go the way of GitHub and allow for unlimited runners since it is in our own infrastructure anyways.

Pro customers are allowed 1 Hybrid runner. The size of the box is defined by your infrastructure. We’ll provide the runner you can use and the concurrency on your account will apply to this runner as well.

The runners were only available for Enterprise customers till now, but being made available to Pro customers at this time as an introduction to the option.

Please tell me if I’ve answered your question

The point is that if it runs on our infrastructure, there shouldn’t be a limit

This is an issue we had when we signed up for the enterprise plan, but didn’t really have other options at the time.

Unless I am misunderstanding this.

For example, we pay for 5 concurrency. However, we should have unlimited given that it is on our infrastructure.

I can take this feedback to our Sales team, but we do not plan to make it unlimited at this time.

I figured as much. GitHub and GitLab definitely went the right way with this.

Sure. I’m just going to be spitting out a lot of information in this thread.

Still no way to change the runtime memory and CPU in the UI

No documentation on how to change the memory and CPU via the CLI

Documentation is under Enterprise, though this is no longer an enterprise feature

Documentation still heavily mentions venona, which while it mentions it is legacy, this is the only place that mentions caching and such

No examples on using the installation options

For example, how do we use the kube-node-selector?

What is storage-class-name’s purpose? The way storage is handled is not documented.

No helm chart was provided (how is this ready for enterprise at all?)

The documentation is poorly laid out and hard to read

During the installation a token is printed out but not explained why we need to keep it

Local SSD is mentioned under Storage space, but no documentation on what to do with it

Local SSD and EBS are not space, they are options. So Storage space as a header makes no sense.

@Harrison Heck Have you already installed the runner? Is this something you would like to do with help at this time?

I’ve already installed it, that part isn’t hard.

The problem is there is no documentation on configuring it

So right now it is useless

I’ll take these questions back and see if we can answer them at least to help you configure it

My honest suggestion is ask your customers to review stuff before releasing things. Because in no way shape or form was the runner’s release done professionally.

https://github.com/codefresh-io/venona - look at this for example, it still acts as if 1.0 is not released yet

The actual GH releases are not following semver

If it requires a ticket to understand how to use it, it wasn’t ready for release.

Thanks Harrison! I understand there are quite a few gaps here

I will take this back to the docs team to make sure all this is filled

I may sound pretty frustrated and I am. CodeFresh does not seem ready for prime time. It feels like an alpha product nearly a year after we started using it.

@Harrison Heck hey Harrison, just wanted to let you know that we’re currently revamping the docs to simplify them. We’ve built up a lot of examples and different areas overtime which makes it confusing to go through.

While there should be some immediate improvements, the larger revamp is underway and will hopefully be ready soon. Thanks for the feedback.

Hey, while that makes sense for other docs, this is documentation for a newly released feature. There isn’t any justification for having launched something new with docs that don’t meet standards.

We were back online a few mins ago, but we are investigating the issue