The OSS Data Synchronization component synchronizes text from an Object Storage Service (OSS) bucket to a MaxCompute table.

Machine Learning Studio allows you to configure the OSS Data Synchronization component only in the Machine Learning Platform for AI (PAI) console. The following table describes the parameters that are used to configure the component.
Parameter Description
OSS Endpoint The endpoint of OSS.
OSS AccessId The AccessKey ID of your Alibaba Cloud account that has OSS activated.
OSS AccessKey The AccessKey secret of your Alibaba Cloud account that has OSS activated.
Bucket The name of your OSS bucket.
Object The OSS object that you want to synchronize.
Column Mapping The mappings between fields in the OSS object and those in the MaxCompute table. Set this parameter in the Index:Name format. Index indicates the index number of the field in the OSS object, and Name indicates the name of the field in the MaxCompute table. Separate multiple mappings with commas (,). Example: 0:label,1:s_width,2:s_length,3:v_width,4:v_length.
Text Delimiter The delimiter that is used to separate the text in the OSS object. Commas (,) are used by default.
Data Compression Format Valid values: No Compression, GZIP, ZIP, and BZIP2.
Text Encoding Format The encoding format of the OSS text. Only UTF-8 is supported.
Data Synchronization Bandwidth Default value: 1. Unit: MB/s.
Stale Data Threshold The maximum number of dirty data records that are allowed. Default value: 0.