Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

initial pass at arch diagrams both abstract and implementation #1245

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

CJ-Wright
Copy link
Member

The idea is to capture all the moving pieces at both in an abstract,
service oriented way and from an implementation perspective.

CC: @beckermr @kruttik-lab49 @k504866430

@CJ-Wright CJ-Wright requested a review from a team as a code owner January 14, 2021 21:25
@kruttik-lab49
Copy link

It might be helpful to add numbered labels to arrows- e.g

staged_recipes >> Edge(label="1. build") >> cis 

This ofcourse has a downside in that it implies a certain directionality.

Overall I think the numbered arrows might be useful.

create_feedstock = GithubActions(label='create feedstock repo')
autotick = GithubActions(label='autotick bot')

anaconda_server = Storage(label='Anaconda Servers')
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not sure if this is too much detail but I think this should contain more detail about how the channel is implemented. Perhaps three components:

  • anaconda.org
  • CDN mirroring with repodata patches being an input
  • The publicly visible channel

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is all stuff anaconda maintains and so we shouldn't have it here. To us anaconda.org is a free service with a big latency for uploads.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it's useful to be here so someone can look at this and understand what is might be wrong with conda-forge for a given symptom. Questions and issues with the CDN sync common enough that I think it's worthwhile acknowledging it exists in the conda-forge documentation. Again, I don't feel that strongly about it if others disagree.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

sure - we can put a big anaconda box around things

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm actually not certain of the anaconda structure. I'm happy to put it in if someone can tell me what the pieces are and how they connect (on the implementation diagram, I think as a concept the channel is just a sink for all of our stuff and a source for users (and ourselves I guess))


staged_recipes >> cis >> staged_recipes
staged_recipes >> linter >> staged_recipes
staged_recipes >> create_feedstock >> new_feedstock >> cis >> validation >> anaconda_server
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not sure if package validation needed for the conceptual diagram as I see it as more of an implementation detail. Perhaps "Anaconda Servers" should be "conda channel"?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We maintain the validation with infrastructure we control so it needs to be here.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I agree it needs to be the "Conda-Forge Implementation" diagram but I'm not sure it needs to be in the "Conda-Forge Conceptual" diagram though I don't feel that strongly about it if others disagree.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ah sure agreed!

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hmm, validation might be an independent concept, since it isn't formally owned by any of the other pieces. Also the fact that we validate our packages is important to the running of conda-forge, since is a barrier to some potential attacks. I would call the implementation detail that this is run off of the heroku server for instance. I'm happy to be turned around on this though.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Development

Successfully merging this pull request may close these issues.

5 participants