TL;DR: Install and enable the volume plugin first (`docker plugin install rexray/ebs`), then create the volume. The daemon cannot use a driver it does not have; the name in your volume or run command must match an installed, enabled plugin.

## The error

```text
Error response from daemon: error looking up volume plugin "rexray": plugin not found
```

## Fix it

1. List installed plugins:
   `docker plugin ls`
   Expected: the driver missing or DISABLED.
2. Install it:
   `docker plugin install rexray/ebs --grant-all-permissions`
   Expected: installs and enables.
3. If present but disabled:
   `docker plugin enable rexray/ebs`
4. Retry the volume create / container run.

## When this applies
- Cloud volume drivers (rexray, portworx), NFS or Ceph plugins referenced but not installed
- Compose files with `driver: rexray` moved between hosts

## When this does NOT apply
- Local driver volumes (built in, never "not found")
- Plugin installed but failing at runtime (check `docker plugin inspect` and plugin logs)

## Versions
Docker Engine 17.06+ (managed plugin system).

## Why it happens
Volume drivers are separate plugin processes. The name is just a string until the daemon resolves it against installed plugins; a typo or a fresh host without the install step fails lookup.

## Edge cases
- Plugins need `--grant-all-permissions` at install or they sit disabled awaiting approval.
- Plugin names include the repo: `docker plugin install vieux/sshfs` then reference driver `vieux/sshfs`.
- After daemon upgrade, plugins sometimes stay disabled; re-enable them.