Branch 1 queue - aajisaka/hadoop-metrics GitHub Wiki
- maps_launched
- Total number of map tasks launched in the queue
- maps_completed
- Total number of map tasks completed in the queue
- maps_failed
- Total number of map tasks failed/killed in the queue
- reduces_launched
- Total number of reduce tasks launched in the queue
- reduces_completed
- Total number of reduce tasks completed in the queue
- reduces_failed
- Total number of reduce tasks failed/killed in the queue
- jobs_submitted
- Total number of jobs submitted in the queue
- jobs_completed
- Total number of jobs completed in the queue
- waiting_maps
- Current number of map tasks waiting in the queue
- waiting_reduces
- Current number of reduce tasks waiting in the queue
- reserved_map_slots
- Current number of map slots reserved in the queue
- reserved_reduce_slots
- Current number of reduce slots reserved in the queue
- jobs_failed
- Total number of failed jobs in the queue
- jobs_killed
- Total number of killed jobs in the queue
- jobs_preparing
- Current number of preparing jobs in the queue
- jobs_running
- Current number of running jobs in the queue
- maps_killed
- 0 (Nothing to increment this metrics)
- reduces_killed
- 0 (Nothing to increment this metrics)
- running_0 (*)
- Current number of running jobs elapsed time are less than 60 minutes
- running_60 (*)
- Current number of running jobs elapsed time are between 60 and 300 minutes
- running_300 (*)
- Current number of running jobs elapsed time are between 300 and 1440 minutes
- running_1440 (*)
- Current number of running jobs elapsed time are more than 1440 minutes
(*) You can set the property mapred.queue.metrics.runtime.buckets in mapred-site.xml to change the buckets (default: 60,300,1440).
For more details, see org.apache.hadoop.mapred.QueueMetrics.java.