flexblock detach

Synopsis

flexblock detach <DEVICE> [flags]

Description

detach gives up ownership of a volume and clears this host’s local state. The volume’s data in object storage is left intact, unlike wipe. Afterward the volume is unowned, so the next host runs a plain attach.

The device must be cleanly stopped, and its cache must hold no dirty slots or pending discards, because that data exists nowhere else. If detach refuses, run start, let the cache flush, stop cleanly, and run detach again.

detach does not need a current grant.

Flags

Flag	Default	Description
--allow-dirty-loss	—	Detach even though dirty slots or an unclean stop mean local data is lost.
--dry-run	—	Print what would happen and exit.

Data loss

--allow-dirty-loss discards data that has not yet reached object storage. Prefer start followed by a clean stop, which flushes it. Use the flag only for a cache that cannot be flushed, such as one superseded by a take-over.

Examples

Detach after a clean stop, before moving the volume to another host:

sudo umount /mnt/analytics
sudo kill -INT "$(pidof flexblock)"
sudo flexblock detach /dev/nvme1n1p1

Give up a cache that another host has taken over (it refuses otherwise with exit code 13):

sudo flexblock detach /dev/nvme1n1p1 --allow-dirty-loss

Related

• flexblock attach
• Moving a Volume
• Stopping Cleanly
• Where Data Can Be Lost