[#559] Add Tree service docs

Signed-off-by: Angira Kekteeva <kira@nspcc.ru>
This commit is contained in:
Angira Kekteeva 2022-06-28 10:05:51 +03:00 committed by Alex Vanin
parent dca7b5bc36
commit 5c8a927c36
2 changed files with 19 additions and 0 deletions

View file

@ -88,6 +88,7 @@ AmazonS3 s3Client = AmazonS3ClientBuilder.standard()
- [Configuration](./docs/configuration.md) - [Configuration](./docs/configuration.md)
- [NeoFS S3 AuthMate](./docs/authmate.md) - [NeoFS S3 AuthMate](./docs/authmate.md)
- [NeoFS Tree service](./docs/tree_service.md)
- [AWS CLI basic usage](./docs/aws_cli.md) - [AWS CLI basic usage](./docs/aws_cli.md)
- [AWS S3 API compatibility](./docs/aws_s3_compat.md) - [AWS S3 API compatibility](./docs/aws_s3_compat.md)
- [AWS S3 Compatibility test results](./docs/s3_test_results.md) - [AWS S3 Compatibility test results](./docs/s3_test_results.md)

18
docs/tree_service.md Normal file
View file

@ -0,0 +1,18 @@
# Tree service
To get objects' metadata and system information, the S3 GW makes requests to the Tree service.
This is a service in NeoFS storage that keeps different information as a tree structure.
Each node keeps one of the types of data as a set of **key-value pairs**:
* Bucket settings: lock configuration and versioning mode
* Bucket tagging
* Object tagging
* Object metadata: OID, name, creation time, system metadata
* Object locking settings
* Active multipart upload info
Some data takes up a lot of memory, so we store it in NeoFS nodes as an object with payload.
But we keep these objects' metadata in the Tree service too:
* Notification configuration
* CORS
* Metadata of parts of active multipart uploads