ZFS pool recovery

A ZFS pool that will not import should be examined before any irreversible rewind.

OpenZFS and TrueNAS pools can fail to import after missing devices, controller or cabling changes, damaged labels, metadata corruption, interrupted writes or insufficient redundancy. Import flags can alter the point in time the pool presents.

Start a recovery case Call 682-325-7712

Understand the symptom

ZFS pools depend on complete vdev topology, replicated metadata and transaction-group consistency.

WHY

What is happening

A pool can contain mirrors, RAIDZ vdevs, special devices, logs, caches and multiple top-level vdevs. Losing a required top-level vdev is different from losing a cache device. Import recovery options may roll the pool back to an earlier transaction group, intentionally discarding newer writes. Device-name changes can also make an intact pool appear missing. Recovery must map the original topology, identify which devices contain required data and preserve labels and uberblock history.

What you may see

Symptoms that change the recovery path.

Pool listed but cannot import

ZFS can see labels while reporting missing devices, I/O errors or corrupt metadata.

Pool disappeared after hardware move

HBA, cabling, device names, sector presentation or enclosures may have changed.

Import suggests -F recovery

ZFS believes rollback to an earlier transaction group may succeed with acknowledged data loss.

One vdev is unavailable

A missing top-level data vdev can prevent the whole pool from importing even when other vdevs are healthy.

Protect the source

Actions that can make the situation worse.

Forced import can change state

Some recovery options roll back transaction groups or write labels and should not be the first action on the only copy.

Replacing the wrong device

Confusing cache, log, special and data vdev roles can create new writes or remove required topology.

Scrub under unstable hardware

A scrub reads the full pool and can expose additional drive failures before data is protected.

Do this now

Safe steps before evaluation.

Record read-only status output

Capture pool name and GUID, zpool import output, zpool status, device identifiers and exact errors.

Map the full vdev topology

Include mirrors, RAIDZ, special, log, cache, spare and expansion devices.

Label physical devices

Preserve bays, serials, HBA ports, enclosures and any device-name changes.

Stop retries and rewinds

Do not repeat force flags, replacement, detach, clear or import recovery on originals.

NEPFOSYS process

How a data-first evaluation proceeds.

Acquire required members

Devices are evaluated and imaged according to their actual ZFS role and condition.

Reconstruct topology and state

Labels, vdev relationships, transaction groups and available recovery points are assessed on protected copies.

Extract and validate datasets

Priority datasets, snapshots, shares and application files are recovered without treating a successful import as the only validation.

A clear next step

Preserve the original device before another repair or recovery attempt.

Start a case with the exact symptoms, incident timeline, prior attempts, encryption details, and the files that matter most. NEPFOSYS will route the case to the appropriate DFW intake or nationwide mail-in path.

Start with a case review

Questions, answered

What customers ask before starting.

Should I run zpool import -F?

Not on the only original pool without understanding the consequence. The option can rewind to an earlier transaction group and discard newer changes. Preserve and test first.

Can a missing cache device prevent recovery?

A cache device is different from a required data or special vdev. The exact topology must be identified before deciding which missing device matters.

Why did the pool disappear after moving disks?

Device paths, HBA presentation, sector size, cabling or missing members may have changed. Preserve labels and use stable device identifiers during evaluation.

DFW local · Nationwide mail-in

Protect the device. Start with the facts.

Start my recovery case