Avoid the massive end-to-end tax of default full history clones

Sicherheitsmeldung mit Schweregrad noch nicht bewertet. Technische Details im Tab „Originaltext“; empfohlene Schritte in der Checkliste.

Originaltext der Quelle (nicht übersetzt) – die deutsche Zusammenfassung findest du im Tab „Kurzfassung".

It's easy to think of git clone as a client-side operation, but the settings of this operation impact the server side and all networks in between. When you run a default full history clone, the server has to walk the entire history, build a pack file for it (that's what "counting objects" is actually doing), and ship it over the wire. The client then unpacks all of it and checks out a full working tree. Every layer, client CPU, network, the Git server's pack-building compute, and disk, pays for the size of that request. Make the request smaller and the whole stack gets cheaper at once, not just your laptop.Agentic AI turns up the pressure on this in a way normal developer workflows don't. An agent doing repository work can clone far more often, and far less predictably, than a human ever would. If the default is a full history clone, you're now paying that tax at agent scale.GitLab is working hard on the backend to serve large repositories faster. But you can also change what you ask for. Options like shallow clones (--depth=1), single-branch clones, and partial clones (--filter=blob:none) let you fetch only what a job actually needs. A more precise request immediately reduces peak load, unreliability, wait time, and cost. And every future backend improvement compounds on top of these leaner requests.In Supercharge your Git workflows, I walked through how Git Much Faster benchmarks the settings that cut clone times by up to 93% and disk usage by up to 98%: disabling compression, widening the HTTP buffer, shallow and partial clones, and sparse checkout that skips binaries. …

Vollständiger Originalartikel ↗

• Betroffenheit im eigenen Stack prüfen: Versionen/Komponenten abgleichen.
• Originalquelle / Hersteller-Advisory lesen: https://about.gitlab.com/blog/git-clone-override-policy/
• Verfügbaren Patch oder Workaround einspielen und dokumentieren.