job claims deis-controller deployment does not exist

Env:
helm/tiller version v2.1.3,
k8s client/master 1.4.6/1.4.7 on GKE,
helmc workflow v2.9.0 install attempting to migrate to helm v2.10.0 (pre-release/staged version) with off-cluster obj storage (gcs) and registry (gcr)

From following steps in `README.md`, after `helm install workflow-migration/workflow-migration`:

```
$ k logs workflow-migration-h5zx7
2017/01/05 22:12:00 Failed to get values: deployments.extensions "deis-controller" not found
```
Even if I re-create it from the saved .yaml,
```
$ k create -f ~/active-deis-controller-deployment.yaml
deployment "deis-controller" created
```
the same error as above occurs, even if the previous job was deleted and created anew.

Not seeing anything super helpful in tiller-deploy logs but here they are just in case:
```
2017/01/05 22:18:54 storage.go:37: Getting release "manageable-stoat" (v1) from storage
2017/01/05 22:18:54 release_server.go:872: Executing pre-install hooks for manageable-stoat
2017/01/05 22:18:54 release_server.go:899: Hooks complete for pre-install manageable-stoat
2017/01/05 22:18:54 storage.go:133: Getting release history for ''
2017/01/05 22:18:54 release_server.go:872: Executing post-install hooks for manageable-stoat
2017/01/05 22:18:54 release_server.go:899: Hooks complete for post-install manageable-stoat
2017/01/05 22:18:54 storage.go:45: Create release "manageable-stoat" (v1) in storage
2017/01/05 22:18:54 storage.go:133: Getting release history for 'manageable-stoat'
2017/01/05 22:18:54 client.go:110: Doing get for: 'workflow-migration'
```