Update RAFT checkpoint when doing a clean shutdown.

[martinmr]

This will prevent the write ahead log from being replayed after the restart of the server.

---

This change is 

[manishrjain]

Reviewable status: 0 of 2 files reviewed, 1 unresolved discussion (waiting on @manishrjain and @martinmr)

---

dgraph/cmd/alpha/run.go, line 509 at r1 (raw file):

		// Update checkpoint so that proposals are not replayed after the server restarts.
		if err := worker.UpdateRaftProgress(); err != nil {

This should be moved to worker.BlockingStop in worker.go.

[martinmr]

Reviewable status: 0 of 2 files reviewed, 1 unresolved discussion (waiting on @manishrjain)

---

dgraph/cmd/alpha/run.go, line 509 at r1 (raw file):

Previously, manishrjain (Manish R Jain) wrote…

This should be moved to worker.BlockingStop in worker.go.

Done.

[manishrjain]

Got a comment.

Reviewed 2 of 3 files at r2.
Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @manishrjain and @martinmr)

---

worker/worker.go, line 114 at r2 (raw file):

	// Update checkpoint so that proposals are not replayed after the server restarts.
	glog.Infof("Updating RAFT state before shutting down...")
	if err := groups().Node.updateRaftProgress(); err != nil {

Might check and see if we need to move this after groups.closer (still keep it before node closer).