Export - CSV (All fields)
Export - CSV (Current fields)
SPARK-11163
Remove unnecessary addPendingTask calls in TaskSetManager.executorLost
Specify number of failed executors in ApplicationMaster error messageSPARK-14468
Always enable OutputCommitCoordinator
hour/minute/second returns negative value
orderBy with multiple columns in WindowSpec does not work properly
Error initializing SparkContext. java.net.URISyntaxException
SQL option "spark.sql.hive.version" doesn't show up in the result of "SET -v"
@Override is not supported by older version of Janino
Create wrong decimal if unscaled > 1e18 and scale > 0
__gettitem__ method throws IndexError exception when we try to access index after the last non-zero entry.
Assertions fail straightforward DataFrame job due to word alignment
[1.5] Table cache for Parquet broken in 1.5
glm does not work with long formula
Elide stacktraces in bin/spark-sql for AnalysisExceptions
StreamContext.getOrCreate is broken is yarn-client mode
cosh is not registered in function registry
Port two minor changes to release packaging scripts back into Spark repo
When killing executor, the pending replacement executors will be lost
PySpark StreamingLogisticRegressionWithSGD does not train with given regParam and convergenceTol parameters
Scan DataSource with predicate expression combine partition key and attributes doesn't work
Tasks that fail due to CommitDeniedException (a side-effect of speculation) can cause job failure
NullPointerException when action called on localCheckpointed RDD (that was checkpointed before)
Internal accumulators miss the internal flag when replaying events in the history server
Flaky test: StandaloneDynamicAllocationSuite
Multivariate Gaussian Model with Covariance matrix returns incorrect answer in some cases
Guard against MAPREDUCE-5918 by ensuring RecordReader is only closed once in *HadoopRDD
A memory leak in SQLListener._stageIdToStageMetrics
Predicates pushed to InmemoryColumnarTableScan are not evaluated correctly
[1.5] Can't create UDF's using hive thrift service
Turns off Parquet filter push-down for string and binary columns
Failure to resolve having correctly
Exchange sort-planning logic incorrectly avoid sorts when existing ordering is non-empty subset of required ordering
RowNumber in HiveContext returns negative values in cluster mode
"ADD JAR" doesn't work if the given path contains URL scheme like "file:/" and "hdfs:/"
Parquet filters push-down may cause exception when schema merging is turned on
Page size calculation is wrong in local mode