聊聊流式数据湖Paimon(四)

--FlinkSQL参数设置
set
    `table.dynamic-table-options.enabled` = `true`;

SET
    `env.state.backend` = `rocksdb`;

SET
    `execution.checkpointing.interval` = `60000`;

SET
    `execution.checkpointing.tolerable-failed-checkpoints` = `3`;

SET
    `execution.checkpointing.min-pause` = `60000`;

--创建Paimon catalog
CREATE CATALOG paimon WITH (
    'type' = 'paimon',
    'metastore' = 'hive',
    'uri' = 'thrift://localhost:9083',
    'warehouse' = 'hdfs://paimon',
    'table.type' = 'EXTERNAL'
);

--创建Partial update结果表
CREATE TABLE if not EXISTS paimon.dw.order_detail (
    `order_id` string,
    `product_type` string,
    `plat_name` string,
    `ref_id` bigint,
    `start_city_name` string,
    `end_city_name` string,
    `create_time` timestamp(3),
    `update_time` timestamp(3),
    `dispatch_time` timestamp(3),
    `decision_time` timestamp(3),
    `finish_time` timestamp(3),
    `order_status` int,
    `binlog_time` bigint,
    PRIMARY KEY (order_id) NOT ENFORCED
) WITH (
    'bucket' = '20',
    -- 指定20个bucket
    'bucket-key' = 'order_id',
    -- 记录排序字段
    'sequence.field' = 'binlog_time',
    -- 选择 full-compaction ，在compaction后产生完整的changelog
    'changelog-producer' = 'full-compaction',
    -- compaction 间隔时间
    'changelog-producer.compaction-interval' = '2 min',
    'merge-engine' = 'partial-update',
    -- 忽略DELETE数据，避免运行报错
    'partial-update.ignore-delete' = 'true'
);

INSERT INTO
    paimon.dw.order_detail 
-- order_info表提供主要字段
SELECT
    order_id,
    product_type,
    plat_name,
    ref_id,
    cast(null as string) as start_city_name,
    cast(null as string) as end_city_name,
    create_time,
    update_time,
    dispatch_time,
    decision_time,
    finish_time,
    order_status,
    binlog_time
FROM
    paimon.ods.order_info
    /*+ OPTIONS ('scan.mode'='latest') */
union
all 
-- order_address表提供城市字段
SELECT
    order_id,
    cast(null as string) as product_type,
    cast(null as string) as plat_name,
    cast(null as bigint) as ref_id,
    start_city_name,
    end_city_name,
    cast(null as timestamp(3)) as create_time,
    cast(null as timestamp(3)) as update_time,
    cast(null as timestamp(3)) as dispatch_time,
    cast(null as timestamp(3)) as decision_time,
    cast(null as timestamp(3)) as finish_time,
    cast(null as int) as order_status,
    binlog_time
FROM
    paimon.ods.order_address
    /*+ OPTIONS ('scan.mode'='latest') */
;

public void testSequenceGroup() {
    sql(
            "CREATE TABLE SG ("
                    + "k INT, a INT, b INT, g_1 INT, c INT, d INT, g_2 INT, PRIMARY KEY (k) NOT ENFORCED)"
                    + " WITH ("
                    + "'merge-engine'='partial-update', "
                    + "'fields.g_1.sequence-group'='a,b', "
                    + "'fields.g_2.sequence-group'='c,d');");

    sql("INSERT INTO SG VALUES (1, 1, 1, 1, 1, 1, 1)");

    // g_2 should not be updated
    sql("INSERT INTO SG VALUES (1, 2, 2, 2, 2, 2, CAST(NULL AS INT))");

    // select *
    assertThat(sql("SELECT * FROM SG")).containsExactlyInAnyOrder(Row.of(1, 2, 2, 2, 1, 1, 1));

    // projection
    assertThat(sql("SELECT c, d FROM SG")).containsExactlyInAnyOrder(Row.of(1, 1));

    // g_1 should not be updated
    sql("INSERT INTO SG VALUES (1, 3, 3, 1, 3, 3, 3)");

    assertThat(sql("SELECT * FROM SG")).containsExactlyInAnyOrder(Row.of(1, 2, 2, 2, 3, 3, 3));

    // d should be updated by null
    sql("INSERT INTO SG VALUES (1, 3, 3, 3, 2, 2, CAST(NULL AS INT))");
    sql("INSERT INTO SG VALUES (1, 4, 4, 4, 2, 2, CAST(NULL AS INT))");
    sql("INSERT INTO SG VALUES (1, 5, 5, 3, 5, CAST(NULL AS INT), 4)");

    assertThat(sql("SELECT a, b FROM SG")).containsExactlyInAnyOrder(Row.of(4, 4));
    assertThat(sql("SELECT c, d FROM SG")).containsExactlyInAnyOrder(Row.of(5, null));
}