1VMTool snapshot machine disk APFS clone, copy नहीं। Filesystem blocks share until write, snapshot zero bytes copy ~millisecond — 24 GiB और 1 TiB same, scale size नहीं।

Snapshots purpose change। Seconds gigabytes cost ration, risky step skip। Nothing cost stop deciding।

Taking और restoring

1vm snapshot take dev "before upgrade"
1vm snapshot list dev
1vm snapshot restore dev before-upgrade
1vm snapshot rm dev <ref>

<ref> name या id prefix। snapshot rm -r whole branch।

Snapshots branch

Older restore new snapshot second line, first discard नहीं। Multiple directions same start, tree view real shape unique data per node — disk consuming branch visible guess नहीं।

Captured

Situation	Captured
Running on Virtualization.framework	Memory disk alongside, restore mid-execution
Running on the 1VM engine	Disk only — cold-boot
Stopped, either engine	Disk only

Three ways back

Confuse easily, plainly stated।

snapshot restore chosen snapshot।

discard-session boot snapshot, power-on since discard:

1vm discard-session dev --force

reset clean state:

1vm reset dev --force

Rollback mode

Never accumulate state — browsing VM, sandbox, demo — rollback on, every write power-off discard:

1vm rollback dev on

Every boot safe point, nothing remember। Structural hand discipline।

Clones

Same clone primitive cheap whole machine copies:

1vm clone dev test            # linked clone, shares blocks until they diverge
1vm clone dev test --full     # an independent copy
1vm clone dev test --rekey-ssh  # regenerate the guest's host keys

Parent alongside --rekey-ssh — same SSH host key otherwise।

Disk usage

1vm du

Allocated, on-disk, per-snapshot bytes library — honest "how much space" clones share blocks।