Creating Iceberg Tables

Creating BookNest's Iceberg Tables

The order history becomes two tables in the booknest namespace, mirroring Analytical SQL and Data Warehouses's orders and order_items. Orders are partitioned with the months transform, giving 18 partitions without a partition column; order lines need no partitioning at this size.

create_tables.sql: BookNest's order tables in the lake catalogSQL

-- BookNest's order tables in the lake catalog (Iceberg format version 2).
CREATE NAMESPACE IF NOT EXISTS booknest;
CREATE TABLE booknest.orders (
  order_id    BIGINT NOT NULL,
  customer_id INT,
  order_ts    TIMESTAMP,             -- Iceberg timestamptz, stored in UTC
  channel     STRING,
  status      STRING,
  coupon      STRING,
  discount    DECIMAL(9,2),
  total       DECIMAL(10,2))
USING iceberg
PARTITIONED BY (months(order_ts))     -- hidden partitioning: no extra column
TBLPROPERTIES ('write.delete.mode' = 'merge-on-read');
CREATE TABLE booknest.order_items (
  order_id   BIGINT NOT NULL,
  line_no    INT,
  book_id    INT,
  qty        INT,
  unit_price DECIMAL(9,2))
USING iceberg;

Spark 129 's TIMESTAMP maps to Iceberg 129 's timestamptz (UTC microseconds), and money stays DECIMAL. The helper demos/ch08/iceberg/run_sql.py runs such files statement by statement with lake.py.

MENU