数据湖最佳实践

SELECT count()
FROM icebergS3('https://my-bucket.s3.amazonaws.com/warehouse/my_table/')
WHERE event_date >= today() - 7

CREATE TABLE events
    ENGINE = IcebergS3('https://my-bucket.s3.amazonaws.com/warehouse/events/')

SELECT count() FROM events WHERE event_date = today()

CREATE DATABASE my_lake
ENGINE = DataLakeCatalog
SETTINGS
    catalog_type = 'glue',
    region = 'us-east-1',
    aws_access_key_id = '<key>',
    aws_secret_access_key = '<secret>'

SELECT count() FROM my_lake.`analytics.events`

SELECT count()
FROM my_lake.`logs.application`
WHERE event_time >= '2026-03-01'
  AND event_time < '2026-03-02'

CREATE TABLE events
    ENGINE = IcebergS3('https://my-bucket.s3.amazonaws.com/warehouse/events/')
SETTINGS iceberg_metadata_async_prefetch_period_ms = 60000;

SELECT count()
FROM events
SETTINGS iceberg_metadata_staleness_ms = 60000;

SELECT count()
FROM my_iceberg_table
SETTINGS iceberg_timestamp_ms = 1714636800000

SET allow_experimental_delta_kernel_rs = 1;

SELECT *
FROM deltaLake('s3://my-bucket/warehouse/ga4_events/')
SETTINGS
    delta_lake_snapshot_start_version = 42,
    delta_lake_snapshot_end_version = 47

SET allow_delta_lake_writes = 1;

INSERT INTO my_delta_table
SETTINGS
    delta_lake_insert_max_rows_in_data_file = 1000000,
    delta_lake_insert_max_bytes_in_data_file = 134217728
SELECT * FROM source_table

CHECK DATABASE my_lake;

SHOW CREATE TABLE my_lake.`db.table`;

SELECT _path, count() AS rows
FROM my_lake.`logs.application`
WHERE event_time >= '2026-03-01'
  AND event_time < '2026-03-02'
GROUP BY _path
ORDER BY rows DESC;

SELECT count() FROM my_iceberg_table
SETTINGS iceberg_metadata_log_level = 'manifest_file_entry';

SYSTEM FLUSH LOGS iceberg_metadata_log;

SELECT content_type, file_path, pruning_status
FROM system.iceberg_metadata_log
WHERE query_id = '<previous_query_id>';