Class SkewAnnotatingPass
- java.lang.Object
-
- org.apache.nemo.common.pass.Pass
-
- org.apache.nemo.compiler.optimizer.pass.compiletime.CompileTimePass
-
- org.apache.nemo.compiler.optimizer.pass.compiletime.annotating.AnnotatingPass
-
- org.apache.nemo.compiler.optimizer.pass.compiletime.annotating.SkewAnnotatingPass
-
public final class SkewAnnotatingPass extends AnnotatingPass
For each shuffle edge, set the number of partitions to (dstParallelism * HASH_RANGE_MULTIPLIER). With this finer-grained partitioning, we can dynamically assign partitions to destination tasks based on data sizes.
-
-
Field Summary
Fields Modifier and Type Field Description static int
HASH_RANGE_MULTIPLIER
Hash range multiplier.
-
Constructor Summary
Constructors Constructor Description SkewAnnotatingPass()
Default constructor.
-
Method Summary
All Methods Instance Methods Concrete Methods Modifier and Type Method Description IRDAG
apply(IRDAG dag)
-
Methods inherited from class org.apache.nemo.compiler.optimizer.pass.compiletime.annotating.AnnotatingPass
getExecutionPropertiesToAnnotate, getPrerequisiteExecutionProperties
-
Methods inherited from class org.apache.nemo.common.pass.Pass
addCondition, getCondition
-
-
-
-
Field Detail
-
HASH_RANGE_MULTIPLIER
public static final int HASH_RANGE_MULTIPLIER
Hash range multiplier. If we need to split or recombine an output data from a task after it is stored, we multiply the hash range with this factor in advance to prevent the extra deserialize - rehash - serialize process. In these cases, the hash range will be (hash range multiplier X destination task parallelism).- See Also:
- Constant Field Values
-
-