Tracking Huge Files with Git LFS
summary
Project originates in Paris on 8 April 2015 and converges Atlassian and GitHub work into Git LFS, a Go implementation that replaces big blobs with pointer files (~100 bytes) and stores content in a parallel LFS store, using sha-256 object IDs. Push uses a pre-push hook to upload objects, pull keeps small pointers and lazily downloads large content during checkout, so a 50 MB image no longer doubles repo size on each edit. Migration uses The BFG to rewrite history; locking and server-side pruning remain evolving features.
AI-written summary · about
chapters (6)
- 0:00Start
- 19:37git filter-branch
- 20:14BFG Repo-Cleaner
- 24:22File locking (single branch model)
- 25:25File locking (multi branch model)
- 34:29Source Tree
related talks
27mGitLFS - How to handle large files in Git
47mMigrating 1TB to GitHub: Lessons Learned from the Field
58mGit at Scale
06mGit Top Tips
22mUsing Git in an (un)orthodox way in your team
21mFighting Git Workflows
31mGit Driven Refactoring
34mSponsored Session: Sustaining Git Performance Under Heavy Workloads