Add more logging to restore state errors

[schmichael]

Sample output:

2017/07/03 11:52:19.185581 [ERR] client: failed to restore state for alloc 9f3a1012-b39b-9494-4d31-0fb4992492fe: failed to read allocation state: failed to read alloc runner alloc_dir state: failed to decode data into passed object: EOF
2017/07/03 11:52:19.185597 [ERR] client: failed to restore state: 1 error(s) occurred:

* failed to read allocation state: failed to read alloc runner alloc_dir state: failed to decode data into passed object: EOF
2017/07/03 11:52:19.185601 [ERR] client: Nomad is unable to start due to corrupt state. The safest way to proceed is to manually stop running task processes and remove Nomad's state dir ("/tmp/nomad-devagent/client") before restarting. Lost allocations will be rescheduled.
2017/07/03 11:52:19.185603 [ERR] client: Corrupt state is often caused by a bug, please report as much information as possible to https://github.com/hashicorp/nomad/issues

To reproduce this error I started a client+server agent (not -dev as dev agents don't save/restore state), ran example.nomad, stopped nomad, and used a little tool to corrupt the state file:

boltq /tmp/nomad-devagent/client/state.db allocations.9f3a1012-b39b-9494-4d31-0fb4992492fe alloc-dir ""

https://github.com/schmichael/boltq

[dadgar]

Probably should wipe both out https://github.com/hashicorp/nomad/blob/master/client/config/config.go#L72-L75

[schmichael]

Good call. Pushed.

[github-actions]

I'm going to lock this pull request because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active contributions.
If you have found a problem that seems related to this change, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.