1.What are the ways to fix the long running jobs in hadoop, any of the jobs (hive/mapr/pig)?
2. A job took 5 mins for a week, but now its taking more than hour. What could be the reason ?
Best place to start will be to check YARN tuning:
There can be number of reasons why your jobs might be running slow like what is the current load on cluster, how is capacity scheduling set up like priority, queues, resources etc.
You can check Hive execution plan to see bottlenecks.