Utils ClassNaiad Help

[This is preliminary documentation and is subject to change.]

Utility classes and methods for the Naiad Storage framework
Inheritance Hierarchy

SystemObject
  Microsoft.Research.Naiad.Frameworks.StorageUtils

Namespace: Microsoft.Research.Naiad.Frameworks.Storage
Assembly: Microsoft.Research.Naiad.Storage (in Microsoft.Research.Naiad.Storage.dll) Version: 0.5.0.0 (0.5.0.0)
Syntax

C#
public static class Utils

The Utils type exposes the following members.

Methods

  NameDescription
Public methodStatic memberDefaultPartFormat(Uri, Int32, Int32, Int32)
default filename generator for a collection partitioned by key, but concatenated across times
Public methodStatic memberDefaultPartFormatTTime(Uri, Int32, Int32, TTime, Int32)
default filename generator for a collection partitioned by key and time
Public methodStatic memberGetNaiadReaderBatchEnumerableTRecord
Enumerates batches of records from a stream in the Naiad serialization format.
Public methodStatic memberGetNaiadReaderEnumerableTRecord
Enumerates records from a stream in the Naiad serialization format.
Public methodStatic memberGetNaiadWriterObserverTRecord
Returns an record observer that writes records to the given stream in the Naiad message format.
Public methodStatic memberReadLines
Enumerates lines of text from a stream
Public methodStatic memberWriteBySubscriptionTOutput, TWriter
serialize a sequence of records to a collection of files partitioned by process and thread. For each process/thread this writes a sequence of files; each time a file reaches a threshold number of bytes, it is closed and another is opened. This keeps individual files of bounded length, allowing for more parallelism when reading them later
Public methodStatic memberWriteByTimeTOutput, TClient, TWriter, TTime
serialize a sequence of records to a collection of files partitioned by process, thread and time. For each process/thread/time this writes a sequence of files; each time a file reaches a threshold number of bytes, it is closed and another is opened. This keeps individual files of bounded length, allowing for more parallelism when reading them later
Top
See Also

Reference