I just switched to using Jekyll 4.3.2 and tried to deploy a static site using this theme, which lists all current GitHub repositories on the generated static site page. To do this, it uses jekyll-github-metadata gem (version 2.16.0) to gather all repositories for my GitHub user.
Apparently, this results in a lot of API requests (for my user, over ~486!) and quickly hits the rate limit when using GITHUB_TOKEN in a GitHub Actions workflow.
Would it be possible to cache or limit API requests further so static site deployments will complete successfully in GitHub Actions?
For example: The jekyll build --verbose output shows each repo is queried for both contributors and releases, but the template doesn't even need this information
Logging at level: debug
Jekyll Version: 4.3.2
Configuration file: /mnt/terabyte/src/sites/trinitronx.github.io/_config.yml
Configuration file: /mnt/terabyte/src/sites/trinitronx.github.io/_config_dev.yml
Requiring: jekyll-github-metadata
To use retry middleware with Faraday v2.0+, install `faraday-retry` gem
GitHub Metadata: Initializing...
Source: /mnt/terabyte/src/sites/trinitronx.github.io
Destination: /mnt/terabyte/src/sites/trinitronx.github.io/_site
Incremental build: enabled
Generating...
Reading: /_layouts/default.html
EntryFilter: excluded /LICENSE.neumorphism
EntryFilter: excluded /README.md
EntryFilter: excluded /docs
EntryFilter: excluded /package.json
EntryFilter: excluded /node_modules
EntryFilter: excluded /.jekyll-cache
EntryFilter: excluded /.github
EntryFilter: excluded /.ruby-version
EntryFilter: excluded /Gemfile
EntryFilter: excluded /yarn.lock
EntryFilter: excluded /.ruby-gemset
EntryFilter: excluded /Gemfile.lock
EntryFilter: excluded /.gitignore
EntryFilter: excluded /gulpfile.js
Reading: index.html
Reading: _data/projects.yml
Reading: _data/skills-frameworks.yml
Reading: _data/skills-languages.yml
Reading: _data/skills-tools.yml
Reading: _data/timeline.yml
Rendering: index.html
Pre-Render Hooks: index.html
Rendering Markup: index.html
Post-Convert Hooks: index.html
Rendering Layout: index.html
GitHub Metadata: Generating for trinitronx/trinitronx.github.io
GitHub Metadata: Calling @client.list_repos("trinitronx", {:type=>"public", :accept=>"application/vnd.github.mercy-preview+json"})
GitHub Metadata: Calling @client.releases("trinitronx/action-doctl")
GitHub Metadata: Calling @client.contributors("trinitronx/action-doctl")
GitHub Metadata: Calling @client.releases("trinitronx/ansible")
GitHub Metadata: Calling @client.contributors("trinitronx/ansible")
GitHub Metadata: Calling @client.releases("trinitronx/ansible-examples")
GitHub Metadata: Calling @client.contributors("trinitronx/ansible-examples")
[... SNIP ...]
[... TOTAL number of releases + contributors API log lines = 486 ...]
[... SNIP ...]
GitHub Metadata: Calling @client.releases("trinitronx/zengge-lightcontrol")
GitHub Metadata: Calling @client.contributors("trinitronx/zengge-lightcontrol")
Writing: /mnt/terabyte/src/sites/trinitronx.github.io/_site/index.html
Writing Metadata: .jekyll-metadata
done in 177.699 seconds.
Auto-regeneration: disabled. Use --watch to enable.
[18:31:02] Finished 'jekyll' after 2.97 min
I just switched to using Jekyll
4.3.2and tried to deploy a static site using this theme, which lists all current GitHub repositories on the generated static site page. To do this, it usesjekyll-github-metadatagem (version2.16.0) to gather all repositories for my GitHub user.Apparently, this results in a lot of API requests (for my user, over ~486!) and quickly hits the rate limit when using
GITHUB_TOKENin a GitHub Actions workflow.Would it be possible to cache or limit API requests further so static site deployments will complete successfully in GitHub Actions?
For example: The
jekyll build --verboseoutput shows each repo is queried for both contributors and releases, but the template doesn't even need this information