Skip to content

try not cleaning up large packages #111

try not cleaning up large packages

try not cleaning up large packages #111

name: mirror-archive-on-merge-to-default-branch
on:
push:
#branches:
# - master
jobs:
mirror-archive:
runs-on: ubuntu-latest
env:
BUCKET: attack-range-attack-data
ATTACK_DATA_ARCHIVE_FILE: attack_data.tar.zstd
steps:
# Use the following GitHub Action to free up disk space on a runner
# This removes dependencies and packages that may be important to many
# GitHub users, such as the Android and dotnet build tools, but are
# not relevant to us
# https://github.com/marketplace/actions/free-disk-space-ubuntu
- name: Free Disk Space (Ubuntu)
uses: jlumbroso/[email protected]
with:
# this might remove tools that are actually needed,
# if set to "true" but frees about 6 GB
tool-cache: false
# all of these default to true, but feel free to set to
# "false" if necessary for your workflow
android: true
dotnet: true
haskell: true
large-packages: false
docker-images: true
swap-storage: true
- name: Checkout Repo
uses: actions/checkout@v4
# We must EXPLICITLY specificy lfs: true. It defaults to false
with:
lfs: true
- name: Setup AWS CLI and Credentials
uses: aws-actions/configure-aws-credentials@v4
with:
aws-access-key-id: ${{ secrets.ACCESS_KEY}}
aws-secret-access-key: ${{ secrets.SECRET_ACCESS_KEY }}
aws-region: us-west-2
- name: Create archive of ONLY the datasets folder
run: |
# The structure of the tar + zstd archive should mirror that of checking out the repo directly
mkdir attack_data
mv datasets/ attack_data/.
#Build some metadata about the archive for documentation purposes
git rev-parse HEAD > attack_data/git_hash.txt
date -u > attack_data/cache_build_date.txt
# Compress with number of threads equal to number of CPU cores.
# Compression level 10 is a great compromise of speed and file size.
# File size reductions are diminishing returns after this - determined experimentally.
tar -c attack_data | zstd --compress -T0 -10 -o $ATTACK_DATA_ARCHIVE_FILE
# - name: Upload Attack data archive file to S3 Bucket
# run: |
# aws s3 cp $ATTACK_DATA_ARCHIVE_FILE s3://$BUCKET/