Skip to main content
Hitachi Vantara Lumada and Pentaho Documentation

S3 File Output

The S3 File Output step writes data as a text file to Amazon Simple Storage Service (S3), a cloud-based storage system.

When you are using Spark as your Adaptive Execution Layer (AEL), the Text File Output step is recommended for writing data to Amazon S3.

Big Data Warning

Because data cannot be appended to S3 using an ‘append mode,’ and the data to send to S3 is generated in memory, an Out Of Memory Exception (OOME) error may occur followed by the “ Read end dead” exception error when the transformation attempts to close the file.

You can use one of the following actions to avoid these errors:

  1. Increase the Java Heap Space (Xmx) setting for Spoon (see Increase the PDI Client Memory Limit).
  2. Set the kettle property s3.vfs.useTempFileOnUploadData=Y

Using the latter method, the system can now generate a temporary file with data and upload it to Amazon S3. In most cases, the user has enough free space to store the temporary file and write permissions on the default temporary-file directory.


Enter the following information in the transformation step name field:

  • Step name: Specifies the unique name of the S3 File Output transformation step on the canvas. You can customize the name or leave it as the default.


The S3 File Output step features several tabs with fields. Each tab is described below. 

File Tab


On the File tab, you can define the basic file properties for the output of this step. 

Option Description

Specify the name of the output text file. The file name of a file in the S3 Cloud uses the following schema:


Do not create file at start  Select to only create the file at the end of the processing cycle. 
Accept file name from field? Select to specify file names in a field in the input stream.
File name field Specify the field that will contain the file names when Accept file name from field is selected.
Extension Specify the three-letter file extension to append to the file name. The default is .txt.
Include stepnr in filename? Select to include the copy number in the file name (_0 for example) when you run the step in multiple copies (launching several copies of a step).
Include partition nr in file name? Select to include the data partition number in the file name.
Include date in file name? Select to include the system date in the filename (_20181231 for example).
Include time in file name? Select to include the system time in the filename (_235959 for example).
Specify Date time format Select to include the date time in the filename using a format from the Date time format dropdown list.
Date time format  Select the date time format. 

Show filename(s)


Click to display a list of the files that will be generated. This is a simulation and depends on the number of rows that go into each file.
Add filenames to result  Clear if you do not want to add file names to the output file.

Content Tab


The Content tab allows you to include additional descriptive fields in the result set.

Option Description
Append Select to append lines to the end of the file.
Separator Specify the character used to separate the fields in a single line of text, typically a semicolon or tab. Click Insert Tab to place a tab in the Separator field. The default value is semicolon (;).
Enclosure Specify to enclose fields with a pair of specified strings. It allows for separator characters in fields. This setting is optional and can be left blank. The default value is double quotes ().
Force the enclosure around fields? Specify to force all field names to be enclosed with the character specified in the Enclosure property.
Header Clear to indicate that the output text file does not have a header row (first line in the file).
Format Specify the type of formatting to use. It can be either DOS or UNIX. UNIX files have lines separated by line feeds, while DOS files have lines separated by carriage returns and line feeds. The default value is CR + LF (Windows, DOS).
Compression Specify the type of compression (.zip or .gzip) to use when compressing the output. Only one file is placed in a single archive. The default value is None.
Encoding Specify the text file encoding to use. Leave blank to use the default encoding on your system. To use Unicode, specify UTF-8 or UTF-16. On first use, PDI searches your system for available encodings.
Right pad fields Select to add spaces to the end of the fields (or remove characters at the end) until the length specified in the table under the Fields tab is reached.
Fast data dump (no formatting) Select to improve the performance when dumping large amounts of data to a text file by not including any formatting information.
Split every ... rows Specify to split the resulting text file into multiple parts of N rows if the number N is larger than zero. 
Add Ending line of file Specify an alternate ending row to the output file.

Fields Tab


The Fields tab allows you to define the properties for the exported fields.

Column Description
Name Specify the name of the field.
Type Select the field's data type from the dropdown list or enter it manually.
Format Select the format mask (number type) from the dropdown list or enter it manually. See Common Formats for information on common valid date and numeric formats you can use in this step.

Specify the length of the field, according to the following field types:

  • Number - Total number of significant figures in a number
  • String - Total length of the string
  • Date - Length of printed output of the string (for example, four is a length for a year)
Precision Specify the number of floating point digits for number-type fields.
Currency Specify the symbol used to represent currencies (for example, $ or ).
Decimal Specify the symbol used to represent a decimal point, either a dot "." or a comma "," (for example, 5,000.00 or 5.000,00).
Group Specify the method used to separate units of thousands in numbers of four digits or larger, either a dot "." or a comma "," (for example, 5,000 or 5.000).
Trim type Select the trimming method to apply to a string, which truncates the field (none, left, right, both) before processing. Trimming only works when no field length is specified.
Null Specify the string to insert into the output text file when the value of the field is null.

Get Fields 


Click to retrieve a list of fields from the input stream.

Minimal width


Click to minimize the field length by removing unnecessary characters. If selected, string fields will no longer be padded to their specified length.

AWS Credentials

The S3 File Output step provides credentials to the Amazon Web Services SDK for Java using a credential provider chain. The default credential  provider chain looks for AWS credentials in the following locations and in the following order:

  • Environment variables – The variables AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, and AWS_SESSION_TOKEN. See AWS Environment Variables
  • AWS credentials file – The credentials file, located in the  /.aws directory on Linux, macOS, and Unix operating systems, and in the "%UserProfile%\.aws directory on Windows operating systems. See AWS Configuration and Credential Files.
  • CLI configuration file – The config file is located in the same directory as the credentials file. The config file can contain a default profile, named profiles, and CLI-specific configuration parameters for each profile.
  • ECS container credentials – These credentials are provided by the Amazon Elastic Container Service on container instances set up by the ECS administrator. See AWS Using an IAM Role to Grant Permissions to Applications.
  • Instance profile credentials – These credentials are delivered through the Amazon EC2 metadata service, and can be used on EC2 instances with an assigned instance role. 

The S3 File Output step can use any of these methods to authenticate AWS credentials. For more information on setting up AWS credentials, see Working with AWS Credentials

Metadata Injection Support

All fields of this step support metadata injection. You can use this step with ETL Metadata Injection to pass metadata to your transformation at runtime.

See Also

S3 CSV Input.