Tigris Import

For Tigris, the S3-compatible API lets SereneDB read and write from Tigris buckets over HTTP(S) directly.

Credentials and Configuration

You will need to generate an access key pair and create an S3 secret in SereneDB:

Query

CREATE SECRET my_secret (    TYPE s3,    KEY_ID '⟨tid_xxxxxxxxxxxx⟩',    SECRET '⟨tsec_xxxxxxxxxxxxxxxxxxxxxxxx⟩',    REGION 'auto',    ENDPOINT 'fly.storage.tigris.dev');

• A single endpoint (fly.storage.tigris.dev) serves all regions; requests are routed to the Tigris region nearest the caller. REGION is required for request signing but is not used for routing — set it to auto.
• URL_STYLE does not need to be set. Tigris uses virtual-hosted-style URLs, which is SereneDB's default for TYPE s3.

Tip

When SereneDB runs on a Fly.io Machine, requests to fly.storage.tigris.dev stay on Fly's internal network and are served from the same region as the Machine when possible.

Querying

After setting up the Tigris credentials, you can query the data using SereneDB's built-in methods, such as read_csv or read_parquet:

Query

SELECT * FROM 's3://⟨tigris-bucket-name⟩/⟨file⟩.csv';
SELECT * FROM read_parquet('s3://⟨tigris-bucket-name⟩/⟨file⟩.parquet');

Result

 id | name  | value----+-------+-------  1 | alpha |    10  2 | beta  |    20
 id | name  | value----+-------+-------  1 | alpha |    10  2 | beta  |    20