59 lines
2.3 KiB
PL/PgSQL
59 lines
2.3 KiB
PL/PgSQL
-- Add migration script here
|
|
CREATE EXTENSION IF NOT EXISTS "uuid-ossp";
|
|
|
|
-- 数据入库记录表
|
|
CREATE TABLE IF NOT EXISTS data_ingestion (
|
|
id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
|
|
ingestion_time TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
|
|
data_time TIMESTAMP NOT NULL,
|
|
source VARCHAR(25) NOT NULL,
|
|
storage_url TEXT NOT NULL,
|
|
created_at TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
|
|
updated_at TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP
|
|
);
|
|
|
|
-- 添加字段注释
|
|
COMMENT ON COLUMN data_ingestion.id IS '数据唯一标识符';
|
|
COMMENT ON COLUMN data_ingestion.ingestion_time IS '入库时间';
|
|
COMMENT ON COLUMN data_ingestion.data_time IS '数据表示的时间';
|
|
COMMENT ON COLUMN data_ingestion.source IS '存放源/数据来源';
|
|
COMMENT ON COLUMN data_ingestion.storage_url IS '存放URL或路径';
|
|
|
|
-- 为快速查找创建索引
|
|
CREATE INDEX IF NOT EXISTS idx_data_ingestion_ingestion_time ON data_ingestion(ingestion_time);
|
|
CREATE INDEX IF NOT EXISTS idx_data_ingestion_id_ingestion_time ON data_ingestion(id, ingestion_time);
|
|
CREATE INDEX IF NOT EXISTS idx_data_ingestion_data_time ON data_ingestion(data_time);
|
|
CREATE INDEX IF NOT EXISTS idx_data_ingestion_source ON data_ingestion(source);
|
|
|
|
-- 可选的复合索引（根据实际查询需求启用）
|
|
-- 如果经常按时间范围+来源查询，取消注释下面的索引
|
|
-- CREATE INDEX IF NOT EXISTS idx_data_ingestion_data_time_source ON data_ingestion(data_time, source);
|
|
-- CREATE INDEX IF NOT EXISTS idx_data_ingestion_source_data_time ON data_ingestion(source, data_time);
|
|
|
|
-- 创建更新时间自动更新的触发器函数
|
|
CREATE OR REPLACE FUNCTION update_updated_at_column()
|
|
RETURNS TRIGGER AS $$
|
|
BEGIN
|
|
NEW.updated_at = CURRENT_TIMESTAMP;
|
|
RETURN NEW;
|
|
END;
|
|
$$ language 'plpgsql';
|
|
|
|
-- 为表创建更新时间触发器
|
|
CREATE TRIGGER update_data_ingestion_updated_at
|
|
BEFORE UPDATE ON data_ingestion
|
|
FOR EACH ROW
|
|
EXECUTE FUNCTION update_updated_at_column();
|
|
|
|
-- 创建用于查询的视图（可选）
|
|
CREATE OR REPLACE VIEW v_data_ingestion_summary AS
|
|
SELECT
|
|
id,
|
|
source,
|
|
COUNT(*) as record_count,
|
|
MIN(data_time) as earliest_data_time,
|
|
MAX(data_time) as latest_data_time,
|
|
MIN(ingestion_time) as first_ingestion_time,
|
|
MAX(ingestion_time) as last_ingestion_time
|
|
FROM data_ingestion
|
|
GROUP BY id, source; |