Linux · Virtualization: QEMU/KVM & libvirt

A VM's image file on the host is far bigger than the data df reports inside the guest. Why, and how do you reclaim the space?

Sign in to see the full answer — free to start on the web.

This is one card from the KnowCard library. The full way to learn it — spaced-repetition review, progress tracking, and AI explanations — lives in the KnowCard app, free to start on the web. iOS & Android are coming soon.

Get started — it’s free

Already have an account? Sign in →

More in Virtualization: QEMU/KVM & libvirt

When you shut down a KVM host, what happens to the running VMs — and where does their state go?
What is the relationship between QEMU and KVM, and why do RHEL and SUSE lean on the pair for enterprise virtualization?
You want to run KVM guests on a cheap cloud VM you rented from a hosting provider. Why does that usually not work?
Beyond ordinary x86 emulation, what makes QEMU able to run software built for a completely different CPU architecture?
KVM is often called a hypervisor, but what does the KVM kernel module actually do — and not do — on its own?
How do you check from the shell whether a CPU supports the hardware virtualization KVM requires, and what does an empty result mean?