UI is incorrectly reporting failed health checks. Using Mesos 0.21.0 with Marathon 0.7.6
1) POST the following to /v2/apps
2) Observe that redis starts up with 3 instances on the mesos cluster. All tasks should be running
3) In the UI, open the task details modal. The health check indicators show red, and hovering over them reports TCP failed health check.
4) Do a query against /v2/apps/redis/tasks - all tasks indicate healthy. All tasks are indeed healthy, and no restarts are taking place against any of the tasks.