Alerting

[Mohit Anchlia]

What's the best way to trigger alert when jobs run for too long or have
many failures? Is there a hadoop command that can be used to perform this
activity?