Replication Resync (both incremental and baseline) Long duration test

[yamingk]

0. Keep I/O running.
1. Restart random member (either leader or follower) -- Randomly enable flip points below to hit either incremental or baseline resync: -- Setup flip point for baseline resync, to simulate the leader and other member have moved ahead with truncation and after the member comes back, it will be a base line resync. -- Setup flip point for the restarted replica to be a fresh member (flip point to simulate a first_time_reboot with homestore format).
2. The restarted member should come back and take log entries from remote and data fetch should be successful.
3. After new member fully synced, it should take incoming I/Os.
4. Stop I/Os on all replicas, and verify I/Os written on each replica.

Repeat above steps for 10 hours.