Export - CSV (All fields)
Export - CSV (Current fields)
MAPREDUCE-4165
Committing is misspelled as commiting in task logs
gridmix has 11 findbugs warnings
RM apps page starttime/endtime sorts are incorrect
Container Logs appear in unsorted order
Move "sleep" and "fail" jobs from tests module to examples moduleMAPREDUCE-4056
Remove MR1 src/test/systemMAPREDUCE-4039
Sort Avoidance
docs need updating on how to start proxyserver
Document the minimum-allocation-mb and maximum-allocation-mb configurations
Sorting by Last Health-Update on the RM nodes page sorts does not work correctly
Error conditions (missing appid, appid not found) are masked in the RM app page
RM capacity scheduler web UI doesn't show active users
Document yarn.nodemanager.delete.debug-delay-sec configuration property
LocalJobRunner limited to single reducer
AM should kill all running tasks on a node when it gets marked as blacklisted
JobHistory is not providing correct count failed,killed task
MR2 reduce tasks showing >100% complete
Make use of fadvise in the NM's shuffle handler
Incorrect custom task status when running on MR2
Add node information to bin/mapred job -list-attempt-ids and other improvements
usercache/<user>/appcache/<appid> directory not removed when using DefaultContainerExecutor
SortedRanges.Range#compareTo is not spec compliant
RM app page shows 500 error on appid parse error
ConcurrentModificationError in DirectoryCollection
ResourceManager throws MetricsException on start up saying QueueMetrics MBean already exists
jobhistory server trying to load job conf file from wrong location
NodeManager status web page should express 'last update' times as seconds ago
If the nodemanager on which the maptask is executed is going down before the mapoutput is consumed by the reducer,then the job is failing with shuffle error
DefaultContainerExecutor cannot launch container under windows
Lower minimum-allocation-mb to sensible defaults
MAPREDUCE-3994
Port TestEmptyJob to maven, possibly using the MiniMRYarnClusterMAPREDUCE-3971
Job History web services need to have limits on the number of items they can return.
MR2 job never completes with 1 pending task
Support fair-sharing option within a MR2 Capacity Scheduler queue
TestYarnClientProtocolProvider is broken
Fix tests in org.apache.hadoop.util.TestRunJar
Fix tests in org.apache.hadoop.ipc.TestSocketFactory
ConcurrentModificationException in FileSystemCounterGroup
Update the HistoryStorage interface to be more complete and robust
Reduce memory usage of counters due to non-static nested classes
YARN should not have a compile-time dependency on HDFS
RM scheduler web page should filter apps to those that are relevant to scheduling
MapReduce should not have a compile-time dependency on HDFS
Fix HA docs for changes to shell command fencer args
Optimising the LineRecordReader initialize() method
Allow symlinking in LocalJobRunner DistributedCache
MR should not be getting duplicate tokens for a MR Job.
Avoid redundant calls for tokens in TokenCache
MR-2450 introduced a significant performance regression (Hive)
Lower default allocation sizes, fix allocation configurations and document them