Attributes: Metadata, Tags, and Labels
You can use attributes to annotate (decorate) your files and trigger pipelines without changing the actual data in your files. There are three different attribute types:
- Metadata
- Tags
- Labels
Each attribute is designed to be used differently. This table provides information describing the different attributes and their usage:
Attribute -> | Metadata | Tag | Label |
|---|---|---|---|
Description | Immutable: Modifying the metadata of a file generates a new file version with a new Attached as native S3 metadata to the underlining S3 object | Immutable: Modifying the tags of a file generates a new file version with a new Attached as native S3 metadata to the underlining S3 object | Mutable: Modifying the labels of a file does not generate a new file version. Files cannot have two labels with the same Files can have two labels with the same name |
Usage | Use to annotate the file with simple information that could be tracked. | Use to store data (for example, the name of a test event). | Use for temporary annotations, or to annotate more complex information. Note: Labels are not archived with TDP, so if you want something more permanent or trackable, consider using metadata or a tag instead. |
Format | key:value pair | String | Array of dictionary(name: a, value: b) |
Sizing | Metadata value: The limit of all metadata and tags associated with a single file is 2 KB. Limited to one key:value | Tag value: The limit of all metadata and tags associated with a single file is 2 KB. Limited to one value. | Label name: 128 characters More than one value is allowed because a label is an array, and one field may store several values. Unlimited number of labels per file. |
Characters Allowed | Keys can include only letters, numbers, spaces and selected symbols ( Values can include only letters, numbers, spaces and selected symbols ( Does not support empty string. | Tags can include only letters, numbers, spaces and selected symbols ( Does not support empty string. | Names can include only letters, numbers, spaces and selected symbols ( Values can include only letters, numbers, spaces and selected symbols ( Does not support empty string. |
Storage | Stored as key:value pair Stored as part of the S3 Object in the S3 Bucket | Stored as part of the S3 Object in the S3 Bucket. | Stored in the database Not stored as part of the S3 Object in the S3 Bucket Not archived with the Tetra Data Platform |
Trigger Pipeline? | Yes | Yes | Yes |
Attribute validation rules apply to all places where you can use metadata, tags, and labels:
- Adding attributes to sources (for example, Tetra Agents)
- Adding attributes to files
- Using the Web API
- Using the ts-sdk
To learn how to manage attributes from the Tetra Data Platform Attribute Management page, click here.
Updated 3 months ago