[ci] switch CRAN mirror to RStudio #5230
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Created based on #5227 (comment).
This project's CI has been blocked for almost a day because cran.r-project.org has been down. According to one of CRAN's maintainers, this is because of an issue with an on-prem apache webserver running in a datacenter in Austria, and it won't be fixed until tomorrow because it's currently the weekend: https://stat.ethz.ch/pipermail/r-devel/2022-May/081750.html.
This PR proposes switching LightGBM's CI to use cran.rstudio.com instead for all its R CI jobs. That mirror is backed by a for-profit company and (I assume) supported by an on-call rotation of engineers who will quickly respond to availability issues. This blog post from 2013 hints how it is architected: https://www.rstudio.com/blog/rstudio-cran-mirror/. This is the default mirror set for everyone using R in the RStudio IDE, so I strongly expect it to have higher availability than cran.r-project.org.
Shouldn't we change links in documentation too?
I intentionally limited the scope of this PR to only LightGBM's CI jobs, and left its documentation pointing to
cran.r-project.org
untouched. I recall CRAN complaining in the past about "non-canonical URLs", and I see the following in https://cran.rstudio.com/doc/manuals/r-release/R-exts.html#Specifying-URLsWriting R Extensions doesn't say whether CRAN is opinionated about which mirror is referenced in such links, but I don't want to risk a rejection by changing the existing README links.
My primary goal with this PR is to unblock LightGBM's development.