ProcessAllInputPartitionsAttribute Class |
Namespace: Ookii.Jumbo.Jet
The ProcessAllInputPartitionsAttribute type exposes the following members.
Name | Description | |
---|---|---|
ProcessAllInputPartitionsAttribute | Initializes a new instance of the ProcessAllInputPartitionsAttribute class |
Name | Description | |
---|---|---|
TypeId | When implemented in a derived class, gets a unique identifier for this Attribute. (Inherited from Attribute.) |
Name | Description | |
---|---|---|
Equals | Returns a value that indicates whether this instance is equal to a specified object. (Inherited from Attribute.) | |
GetHashCode | Returns the hash code for this instance. (Inherited from Attribute.) | |
GetType | Gets the Type of the current instance. (Inherited from Object.) | |
IsDefaultAttribute | When overridden in a derived class, indicates whether the value of this instance is the default value for the derived class. (Inherited from Attribute.) | |
Match | When overridden in a derived class, returns a value that indicates whether this instance equals a specified object. (Inherited from Attribute.) | |
ToString | Returns a string that represents the current object. (Inherited from Object.) |
The only time a task can have multiple input partitions is if its input is a file channel and PartitionsPerTask is larger than 1. This attribute has no effect on child tasks of a compound task with internal partitioning.
If the task that this attribute is applied to is a pull task, it can determine what partition is being processed and how many partitions there are by casting the input record reader to a MultiPartitionRecordReaderT.
However, if the input to a pull task with the ProcessAllInputPartitionsAttribute attribute is not a channel with multiple partitions per task, the input record reader will not be a MultiPartitionRecordReaderT so you should not assume that such a cast will always succeed.
The input partition affinity of the output of a task with the ProcessAllInputPartitionsAttribute attribute is lost. If the output is a child task, it cannot determine what the current input partition is, and if the output is written to the DFS, it will be a single file (rather than a file for each partition, which is the default).