On this Page

Snap type:

Format


Description:

This Snap formats the incoming document from the upstream Snaps to Hadoop sequence file format, the native binary data format to persist intermediate data between different stages of MapReduce jobs. 

To enable Snappy compression for sequence file, a cluster level setting needs to be set. Unlike Parquet or ORC writer the Sequence Formatter Snap will not have "Snappy" listed in the "compression" options.
  • Expected upstream Snaps: The upstream Snap for Sequence Formatter should output map/table/key-value formatted data. Valid data types include String, Integer, Number and Boolean.
  • Expected downstream Snaps: The Sequence Formatter Snap outputs binary data, so the downstream Snap must be a data store output Snap like (File Writer, HDFS Writer, etc.).


Prerequisites:

[None]


Support and limitations:Works in Ultra Pipelines.
Account: 

Accounts are not used with this Snap.


Views:


InputThis Snap has at most one document input view.
OutputThis Snap has at most one binary output view.
Error

This Snap has at most one document error view and produces zero or more documents in the view. 

In Spark mode:

  • the errors will be routed to error view documents if the error policy is defined as CONTINUE;
  • the execution will stop on first error if the error policy is defined as FAIL;
  • the execution will simply ignore the error if the error policy is defined as IGNORE.


Settings

Label


Required. The name for the Snap. You can modify this to be more specific, especially if you have more than one of the same Snap in your pipeline.

Key



Required. JSON path for the key.

Example: $input_column_name

Default value: [None]


Value


Required. JSON path for the value.

Example: $input_column_name

Default value:  [None]


Compression type


For Spark mode, when compression is required, this value has to be populated. In standard mode, it is not required.

Sequence file compression type. The options available include:

  • Record: Only values are compressed.
  • Block: Both keys and values are compressed.
  • [None]: Records are uncompressed when none is selected.

Default value: [None]


Compression codec


Required for any Compression Type in Spark mode.

Fully qualified compression codec class name

Example: org.apache.hadoop.io.compress.GzipCodec


Troubleshooting