client: always mark exited sys/svc allocs as failed

drain: fix draining of system jobs

Create evals for system jobs when drain is unset

This PR creates evals for system jobs when:

* Drain is unset and mark eligible is true
* Eligibility is restored to the node

When restarts.attempts=0 was set in a jobspec a system or service alloc
that exited with 0 status would be marked as `completed` instead of
`failed`. Since system and service jobs are intended to run until
stopped or updated, they should always be marked as failed when they
exit even in cases where the exit code is 0.

May want to bump this a bit more as the drain often completes amidst a
flurry of alloc updates.

Since the BatchJob helper had a different task group than the Alloc
helper, it was difficult to create a valid batch alloc.

System allocs should be drained when a node's deadline is hit or when
all other allocs on the node have stopped/migrated.

Properly format node event timestamps

docs for delayed rescheduling options

Only track nodes if the conn is from the node

Fix alloc watcher snapshot streaming

Correct status desc on draining system allocs

Fixes a bug in which a connection to a Nomad server was treated as a
connection to a node because the server forwarded a node specific RPC.

Allow TLS configurations for HTTP and RPC connections to be reloaded …

s/linear/constant/g

Nomad now closes old connections on reloading a TLS configuration

Adds additional validation for ambiguous settings 

Fix edge case in reconciler 

fix typo and remove duplicated text in Vault stanza docs

Only create deployment if job changes or no allocs