## TL;DR
The target datastore being inaccessible at restore time is a host-side storage condition: check the datastore's state on the target host in the hypervisor, whether it is mounted, its free space, and whether a storage event, a lost LUN or an NFS outage, made it vanish after the job was configured. Pick a healthy target or restore the datastore's connectivity, then re-run the restore; the backup data is not the problem in this failure.

```text
Nakivo restore failed "failed to create target VM: datastore is inaccessible"
```

## Use this when
- The job or restore fails with this symptom: Nakivo restore failed with failed to create target VM: datastore is inaccessible.
- You are working on a Nakivo restore creating a VM on a target datastore and the failure repeats on the same target or schedule.
- You need the failing job, stage, or component isolated before changing retention or infrastructure.

## Not for this skill when
- A datastore that is accessible but out of space, where the restore's error wording still says inaccessible and the fix is capacity.
- The error text names a different product, stage, or root cause. Match the verbatim error first and use that skill.

## Steps
1. Check the target datastore in the hypervisor's view on the intended host: mounted, visible, and with free space. Verify: datastore condition is established
2. Look for storage events on that host in the window: path failures, unmounts, or NFS server issues. Verify: any storage event is correlated
3. Test creating a small placeholder VM or file on the datastore from the hypervisor side. Verify: the datastore accepts new objects, or refuses informatively
4. Restore datastore health or select a different target, then rerun the restore and confirm the VM is created. Verify: the restore creates the target VM

## Variant phrasings
### nakivo restore failed create target vm datastore inaccessible
The verbatim phrasing.
### nakivo restore datastore inaccessible target host fix
The target phrasing.
### nakivo vm restore failed datastore unmounted space
The hypervisor-view phrasing.

Compatibility: Nakivo Backup and Replication restores and target datastores

## Why it happens
Restores validate their target at creation time, so a datastore that was healthy when the job was designed can be a refusal when the restore runs. The message collapses unmounted, unreachable, and sometimes full into one word, and only the hypervisor's own view separates them.

## Edge cases / pitfalls
- Datastores visible on one host may be missing on another in the same cluster; the restore's target host is the view that matters.
- After a storage outage, datastores sometimes remount read-only; a create test catches what a status view does not.

## Provenance

Resolved from the public thread: https://vectle.com/posts/pst_9J_MZSEa3A-GbpJo0pOjpg