jettask 0.2.5__py3-none-any.whl → 0.2.7__py3-none-any.whl

jettask/webui/sql/init_database.sql

@@ -0,0 +1,640 @@
+-- JetTask 数据库初始化脚本
+-- 包含所有表结构、索引、函数和分区设置
+
+-- ========================================
+-- 1. 创建更新时间触发器函数
+-- ========================================
+CREATE OR REPLACE FUNCTION update_updated_at_column()
+RETURNS TRIGGER AS $$
+BEGIN
+    NEW.updated_at = CURRENT_TIMESTAMP;
+    RETURN NEW;
+END;
+$$ LANGUAGE plpgsql;
+
+-- ========================================
+-- 2. 创建 namespaces 表
+-- ========================================
+CREATE TABLE IF NOT EXISTS namespaces (
+    id SERIAL PRIMARY KEY,
+    name VARCHAR(255) UNIQUE NOT NULL,
+    description TEXT,
+    connection_url TEXT NOT NULL,
+    is_default BOOLEAN DEFAULT FALSE,
+    created_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    updated_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP
+);
+
+-- 插入默认命名空间
+INSERT INTO namespaces (name, description, connection_url, is_default)
+VALUES ('default', 'Default namespace', 'redis://localhost:6379/0', true)
+ON CONFLICT (name) DO NOTHING;
+
+-- ========================================
+-- 3. 创建 scheduled_tasks 表
+-- ========================================
+CREATE TABLE IF NOT EXISTS scheduled_tasks (
+    id BIGSERIAL PRIMARY KEY,
+    scheduler_id VARCHAR(255) UNIQUE NOT NULL,
+    task_name VARCHAR(255) NOT NULL,
+    task_type VARCHAR(50) NOT NULL,
+    queue_name VARCHAR(100) NOT NULL,
+    task_args JSONB DEFAULT '[]'::jsonb,
+    task_kwargs JSONB DEFAULT '{}'::jsonb,
+    cron_expression VARCHAR(100),
+    interval_seconds NUMERIC(10,2),
+    next_run_time TIMESTAMPTZ,
+    last_run_time TIMESTAMPTZ,
+    enabled BOOLEAN DEFAULT true,
+    max_retries INTEGER DEFAULT 3,
+    retry_delay INTEGER DEFAULT 60,
+    timeout INTEGER DEFAULT 300,
+    description TEXT,
+    tags JSONB DEFAULT '[]'::jsonb,
+    metadata JSONB DEFAULT '{}'::jsonb,
+    created_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    updated_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    execution_count INTEGER DEFAULT 0,
+    namespace VARCHAR(255) DEFAULT 'default',
+    priority INTEGER
+);
+
+-- 创建优化后的索引
+CREATE INDEX IF NOT EXISTS idx_scheduled_tasks_next_run 
+ON scheduled_tasks (next_run_time) 
+INCLUDE (scheduler_id, task_name, queue_name, enabled)
+WHERE enabled = true;
+
+CREATE INDEX IF NOT EXISTS idx_scheduled_tasks_enabled_ns
+ON scheduled_tasks (namespace, enabled)
+WHERE enabled = true;
+
+-- 创建触发器
+CREATE TRIGGER update_scheduled_tasks_updated_at 
+BEFORE UPDATE ON scheduled_tasks 
+FOR EACH ROW 
+EXECUTE FUNCTION update_updated_at_column();
+
+-- ========================================
+-- 4. 创建 tasks 分区表
+-- ========================================
+CREATE TABLE IF NOT EXISTS tasks (
+    stream_id TEXT NOT NULL,
+    queue TEXT NOT NULL,
+    namespace TEXT NOT NULL DEFAULT 'default',
+    scheduled_task_id TEXT,
+    payload JSONB NOT NULL,
+    priority INTEGER DEFAULT 0,
+    created_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    source TEXT NOT NULL DEFAULT 'redis_stream',
+    metadata JSONB DEFAULT '{}',
+    PRIMARY KEY (stream_id, created_at)
+) PARTITION BY RANGE (created_at);
+
+-- 创建索引
+CREATE INDEX IF NOT EXISTS idx_tasks_created 
+ON tasks (created_at DESC);
+
+CREATE INDEX IF NOT EXISTS idx_tasks_ns_created 
+ON tasks (namespace, created_at DESC)
+INCLUDE (queue, priority);
+
+CREATE INDEX IF NOT EXISTS idx_tasks_queue_time
+ON tasks (queue, created_at DESC)
+WHERE namespace = 'default';
+
+CREATE INDEX IF NOT EXISTS idx_tasks_scheduled_id
+ON tasks (scheduled_task_id, created_at DESC)
+WHERE scheduled_task_id IS NOT NULL;
+
+-- ========================================
+-- 5. 创建 task_runs 分区表
+-- ========================================
+CREATE TABLE IF NOT EXISTS task_runs (
+    id BIGSERIAL,
+    stream_id TEXT NOT NULL,
+    task_name TEXT NOT NULL,
+    consumer_group TEXT NOT NULL,
+    consumer_name TEXT,
+    worker_id TEXT,
+    status TEXT NOT NULL DEFAULT 'pending',
+    created_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    start_time TIMESTAMPTZ,
+    end_time TIMESTAMPTZ,
+    retry_count INTEGER DEFAULT 0,
+    max_retries INTEGER DEFAULT 3,
+    error_message TEXT,
+    error_details JSONB,
+    result JSONB,
+    logs TEXT[],
+    updated_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    duration DOUBLE PRECISION,
+    execution_time DOUBLE PRECISION,
+    PRIMARY KEY (id, created_at),
+    UNIQUE (stream_id, consumer_group, created_at)
+) PARTITION BY RANGE (created_at);
+
+-- 创建索引
+CREATE INDEX IF NOT EXISTS idx_task_runs_stream_covering 
+ON task_runs (stream_id, created_at) 
+INCLUDE (status, execution_time, end_time, start_time);
+
+CREATE INDEX IF NOT EXISTS idx_task_runs_status_created
+ON task_runs (status, created_at);
+
+CREATE INDEX IF NOT EXISTS idx_task_runs_success_time
+ON task_runs (end_time DESC, created_at) 
+WHERE status = 'success';
+
+CREATE INDEX IF NOT EXISTS idx_task_runs_consumer_created 
+ON task_runs (consumer_group, created_at DESC);
+
+-- 创建触发器
+CREATE TRIGGER update_task_runs_updated_at 
+BEFORE UPDATE ON task_runs 
+FOR EACH ROW 
+EXECUTE FUNCTION update_updated_at_column();
+
+-- ========================================
+-- 6. 创建 stream_backlog_monitor 分区表
+-- ========================================
+CREATE TABLE IF NOT EXISTS stream_backlog_monitor (
+    id BIGSERIAL,
+    namespace TEXT NOT NULL DEFAULT 'default',
+    stream_name TEXT NOT NULL,
+    consumer_group TEXT,
+    priority INTEGER,
+    last_published_offset TEXT,
+    last_consumed_offset TEXT,
+    last_delivered_offset BIGINT DEFAULT 0,  -- 最后投递的偏移量
+    last_acked_offset BIGINT DEFAULT 0,      -- 最后确认的偏移量
+    consumer_lag BIGINT,
+    pending_count BIGINT DEFAULT 0,          -- 待处理任务数
+    backlog_undelivered BIGINT DEFAULT 0,    -- 未投递的积压数
+    backlog_unprocessed BIGINT DEFAULT 0,    -- 未处理的积压数
+    task_count BIGINT,
+    created_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    collected_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    metadata JSONB DEFAULT '{}',
+    PRIMARY KEY (id, created_at)
+) PARTITION BY RANGE (created_at);
+
+-- 创建索引
+CREATE INDEX IF NOT EXISTS idx_stream_backlog_namespace_stream 
+ON stream_backlog_monitor (namespace, stream_name, created_at DESC);
+
+CREATE INDEX IF NOT EXISTS idx_stream_backlog_created 
+ON stream_backlog_monitor (created_at DESC);
+
+-- ========================================
+-- 7. 创建 alert_rules 表
+-- ========================================
+CREATE TABLE IF NOT EXISTS alert_rules (
+    id BIGSERIAL PRIMARY KEY,
+    name VARCHAR(255) NOT NULL,
+    description TEXT,
+    rule_type VARCHAR(50) NOT NULL,
+    metric VARCHAR(100) NOT NULL,
+    condition VARCHAR(20) NOT NULL,
+    threshold DOUBLE PRECISION NOT NULL,
+    duration_seconds INTEGER DEFAULT 60,
+    enabled BOOLEAN DEFAULT true,
+    alert_channels JSONB DEFAULT '[]'::jsonb,
+    metadata JSONB DEFAULT '{}'::jsonb,
+    created_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    updated_at TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP
+);
+
+CREATE INDEX IF NOT EXISTS idx_alert_rules_enabled 
+ON alert_rules (enabled) WHERE enabled = true;
+
+-- 创建触发器
+CREATE TRIGGER update_alert_rules_updated_at 
+BEFORE UPDATE ON alert_rules 
+FOR EACH ROW 
+EXECUTE FUNCTION update_updated_at_column();
+
+-- ========================================
+-- 8. 创建 alert_history 表
+-- ========================================
+CREATE TABLE IF NOT EXISTS alert_history (
+    id BIGSERIAL PRIMARY KEY,
+    rule_id BIGINT REFERENCES alert_rules(id),
+    alert_time TIMESTAMPTZ DEFAULT CURRENT_TIMESTAMP,
+    metric_value DOUBLE PRECISION,
+    threshold_value DOUBLE PRECISION,
+    alert_message TEXT,
+    alert_level VARCHAR(20),
+    resolved BOOLEAN DEFAULT false,
+    resolved_time TIMESTAMPTZ,
+    metadata JSONB DEFAULT '{}'::jsonb
+);
+
+CREATE INDEX IF NOT EXISTS idx_alert_history_rule_id 
+ON alert_history (rule_id, alert_time DESC);
+
+CREATE INDEX IF NOT EXISTS idx_alert_history_unresolved 
+ON alert_history (resolved, alert_time DESC) WHERE resolved = false;
+
+-- ========================================
+-- 9. 分区管理函数
+-- ========================================
+
+-- tasks 表分区管理
+CREATE OR REPLACE FUNCTION create_tasks_partition()
+RETURNS void AS $$
+DECLARE
+    start_date date;
+    end_date date;
+    partition_name text;
+BEGIN
+    -- 创建当前月分区
+    start_date := date_trunc('month', CURRENT_DATE);
+    end_date := start_date + interval '1 month';
+    partition_name := 'tasks_' || to_char(start_date, 'YYYY_MM');
+    
+    IF NOT EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = partition_name
+    ) THEN
+        EXECUTE format(
+            'CREATE TABLE %I PARTITION OF tasks 
+            FOR VALUES FROM (%L) TO (%L)',
+            partition_name, start_date, end_date
+        );
+        RAISE NOTICE 'Created partition: %', partition_name;
+    END IF;
+    
+    -- 创建下月分区
+    start_date := date_trunc('month', CURRENT_DATE + interval '1 month');
+    end_date := start_date + interval '1 month';
+    partition_name := 'tasks_' || to_char(start_date, 'YYYY_MM');
+    
+    IF NOT EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = partition_name
+    ) THEN
+        EXECUTE format(
+            'CREATE TABLE %I PARTITION OF tasks 
+            FOR VALUES FROM (%L) TO (%L)',
+            partition_name, start_date, end_date
+        );
+        RAISE NOTICE 'Created partition: %', partition_name;
+    END IF;
+    
+    -- 创建上月分区
+    start_date := date_trunc('month', CURRENT_DATE - interval '1 month');
+    end_date := start_date + interval '1 month';
+    partition_name := 'tasks_' || to_char(start_date, 'YYYY_MM');
+    
+    IF NOT EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = partition_name
+    ) THEN
+        EXECUTE format(
+            'CREATE TABLE %I PARTITION OF tasks 
+            FOR VALUES FROM (%L) TO (%L)',
+            partition_name, start_date, end_date
+        );
+        RAISE NOTICE 'Created partition: %', partition_name;
+    END IF;
+END;
+$$ LANGUAGE plpgsql;
+
+-- task_runs 表分区管理
+CREATE OR REPLACE FUNCTION create_task_runs_partition()
+RETURNS void AS $$
+DECLARE
+    start_date date;
+    end_date date;
+    partition_name text;
+BEGIN
+    -- 创建当前月份分区
+    start_date := date_trunc('month', CURRENT_DATE);
+    end_date := start_date + interval '1 month';
+    partition_name := 'task_runs_' || to_char(start_date, 'YYYY_MM');
+    
+    IF NOT EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = partition_name
+    ) THEN
+        EXECUTE format(
+            'CREATE TABLE %I PARTITION OF task_runs 
+            FOR VALUES FROM (%L) TO (%L)',
+            partition_name, start_date, end_date
+        );
+        RAISE NOTICE 'Created partition: %', partition_name;
+    END IF;
+    
+    -- 创建下个月分区
+    start_date := date_trunc('month', CURRENT_DATE + interval '1 month');
+    end_date := start_date + interval '1 month';
+    partition_name := 'task_runs_' || to_char(start_date, 'YYYY_MM');
+    
+    IF NOT EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = partition_name
+    ) THEN
+        EXECUTE format(
+            'CREATE TABLE %I PARTITION OF task_runs 
+            FOR VALUES FROM (%L) TO (%L)',
+            partition_name, start_date, end_date
+        );
+        RAISE NOTICE 'Created partition: %', partition_name;
+    END IF;
+    
+    -- 创建前一个月分区
+    start_date := date_trunc('month', CURRENT_DATE - interval '1 month');
+    end_date := start_date + interval '1 month';
+    partition_name := 'task_runs_' || to_char(start_date, 'YYYY_MM');
+    
+    IF NOT EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = partition_name
+    ) THEN
+        EXECUTE format(
+            'CREATE TABLE %I PARTITION OF task_runs 
+            FOR VALUES FROM (%L) TO (%L)',
+            partition_name, start_date, end_date
+        );
+        RAISE NOTICE 'Created partition: %', partition_name;
+    END IF;
+END;
+$$ LANGUAGE plpgsql;
+
+-- stream_backlog_monitor 表分区管理
+CREATE OR REPLACE FUNCTION create_stream_backlog_partition()
+RETURNS void AS $$
+DECLARE
+    start_date date;
+    end_date date;
+    partition_name text;
+BEGIN
+    -- 创建最近3个月的分区
+    FOR i IN -1..1 LOOP
+        start_date := date_trunc('month', CURRENT_DATE + (i || ' month')::interval);
+        end_date := start_date + interval '1 month';
+        partition_name := 'stream_backlog_monitor_' || to_char(start_date, 'YYYY_MM');
+        
+        IF NOT EXISTS (
+            SELECT 1 FROM pg_class WHERE relname = partition_name
+        ) THEN
+            EXECUTE format(
+                'CREATE TABLE %I PARTITION OF stream_backlog_monitor 
+                FOR VALUES FROM (%L) TO (%L)',
+                partition_name, start_date, end_date
+            );
+            RAISE NOTICE 'Created partition: %', partition_name;
+        END IF;
+    END LOOP;
+END;
+$$ LANGUAGE plpgsql;
+
+-- ========================================
+-- 10. 分区维护函数
+-- ========================================
+
+-- tasks 表维护
+CREATE OR REPLACE FUNCTION maintain_tasks_partitions()
+RETURNS void AS $$
+DECLARE
+    partition_date date;
+    partition_name text;
+    old_partition_date date;
+    old_partition_name text;
+BEGIN
+    -- 创建未来分区
+    PERFORM create_tasks_partition();
+    
+    -- 删除超过2个月的旧分区
+    old_partition_date := date_trunc('month', CURRENT_DATE - interval '2 months');
+    old_partition_name := 'tasks_' || to_char(old_partition_date, 'YYYY_MM');
+    
+    IF EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = old_partition_name
+    ) THEN
+        EXECUTE format('DROP TABLE IF EXISTS %I CASCADE', old_partition_name);
+        RAISE NOTICE 'Dropped old partition: %', old_partition_name;
+    END IF;
+    
+    -- 对活跃分区执行 VACUUM ANALYZE
+    FOR partition_date IN 
+        SELECT generate_series(
+            date_trunc('month', CURRENT_DATE - interval '1 month'),
+            date_trunc('month', CURRENT_DATE),
+            interval '1 month'
+        )::date
+    LOOP
+        partition_name := 'tasks_' || to_char(partition_date, 'YYYY_MM');
+        IF EXISTS (
+            SELECT 1 FROM pg_class WHERE relname = partition_name
+        ) THEN
+            EXECUTE format('VACUUM ANALYZE %I', partition_name);
+        END IF;
+    END LOOP;
+END;
+$$ LANGUAGE plpgsql;
+
+-- task_runs 表维护
+CREATE OR REPLACE FUNCTION maintain_task_runs_partitions()
+RETURNS void AS $$
+DECLARE
+    partition_date date;
+    partition_name text;
+    old_partition_date date;
+    old_partition_name text;
+BEGIN
+    -- 创建未来分区
+    PERFORM create_task_runs_partition();
+    
+    -- 删除超过3个月的旧分区
+    old_partition_date := date_trunc('month', CURRENT_DATE - interval '3 months');
+    old_partition_name := 'task_runs_' || to_char(old_partition_date, 'YYYY_MM');
+    
+    IF EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = old_partition_name
+    ) THEN
+        EXECUTE format('DROP TABLE IF EXISTS %I', old_partition_name);
+        RAISE NOTICE 'Dropped old partition: %', old_partition_name;
+    END IF;
+    
+    -- 对活跃分区执行 VACUUM ANALYZE
+    FOR partition_date IN 
+        SELECT generate_series(
+            date_trunc('month', CURRENT_DATE - interval '1 month'),
+            date_trunc('month', CURRENT_DATE + interval '1 month'),
+            interval '1 month'
+        )::date
+    LOOP
+        partition_name := 'task_runs_' || to_char(partition_date, 'YYYY_MM');
+        IF EXISTS (
+            SELECT 1 FROM pg_class WHERE relname = partition_name
+        ) THEN
+            EXECUTE format('VACUUM ANALYZE %I', partition_name);
+            RAISE NOTICE 'Vacuumed partition: %', partition_name;
+        END IF;
+    END LOOP;
+END;
+$$ LANGUAGE plpgsql;
+
+-- scheduled_tasks 表维护
+CREATE OR REPLACE FUNCTION maintain_scheduled_tasks()
+RETURNS void AS $$
+BEGIN
+    -- 重置长时间未执行的任务
+    UPDATE scheduled_tasks
+    SET next_run_time = CURRENT_TIMESTAMP
+    WHERE enabled = true
+      AND next_run_time < CURRENT_TIMESTAMP - interval '1 hour';
+    
+    -- 记录维护日志
+    RAISE NOTICE 'Scheduled tasks maintenance completed at %', CURRENT_TIMESTAMP;
+    
+    -- 更新统计信息
+    ANALYZE scheduled_tasks;
+END;
+$$ LANGUAGE plpgsql;
+
+-- ========================================
+-- 11. 创建初始分区
+-- ========================================
+SELECT create_tasks_partition();
+SELECT create_task_runs_partition();
+SELECT create_stream_backlog_partition();
+
+-- ========================================
+-- 12. 批量操作优化函数
+-- ========================================
+
+-- task_runs批量UPSERT函数
+CREATE OR REPLACE FUNCTION batch_upsert_task_runs(
+    p_records jsonb
+) RETURNS INTEGER AS $$
+DECLARE
+    v_count INTEGER := 0;
+    v_record jsonb;
+BEGIN
+    -- 遍历每条记录
+    FOR v_record IN SELECT * FROM jsonb_array_elements(p_records)
+    LOOP
+        -- 先尝试UPDATE
+        UPDATE task_runs SET
+            consumer_name = COALESCE((v_record->>'consumer_name')::TEXT, consumer_name),
+            status = CASE 
+                WHEN (v_record->>'status')::TEXT IS NULL THEN status
+                WHEN status = 'pending' THEN COALESCE((v_record->>'status')::TEXT, status)
+                WHEN status = 'running' AND (v_record->>'status')::TEXT IN ('success', 'failed', 'timeout', 'skipped') THEN (v_record->>'status')::TEXT
+                WHEN status IN ('success', 'failed', 'timeout', 'skipped') THEN status
+                ELSE COALESCE((v_record->>'status')::TEXT, status)
+            END,
+            result = CASE
+                WHEN status IN ('success', 'failed', 'timeout', 'skipped') AND (v_record->>'status')::TEXT NOT IN ('success', 'failed', 'timeout', 'skipped') THEN result
+                ELSE COALESCE((v_record->>'result')::jsonb, result)
+            END,
+            error_message = CASE
+                WHEN status IN ('success', 'failed', 'timeout', 'skipped') AND (v_record->>'status')::TEXT NOT IN ('success', 'failed', 'timeout', 'skipped') THEN error_message
+                ELSE COALESCE((v_record->>'error_message')::TEXT, error_message)
+            END,
+            start_time = COALESCE((v_record->>'started_at')::TIMESTAMPTZ, start_time),
+            end_time = CASE
+                WHEN status IN ('success', 'failed', 'timeout', 'skipped') AND (v_record->>'status')::TEXT NOT IN ('success', 'failed', 'timeout', 'skipped') THEN end_time
+                ELSE COALESCE((v_record->>'completed_at')::TIMESTAMPTZ, end_time)
+            END,
+            worker_id = COALESCE((v_record->>'worker_id')::TEXT, worker_id),
+            duration = COALESCE((v_record->>'duration')::DOUBLE PRECISION, duration),
+            execution_time = COALESCE((v_record->>'execution_time')::DOUBLE PRECISION, execution_time),
+            updated_at = CURRENT_TIMESTAMP
+        WHERE stream_id = (v_record->>'stream_id')::TEXT 
+          AND consumer_group = (v_record->>'consumer_group')::TEXT;
+        
+        -- 如果没有更新到任何行，则INSERT
+        IF NOT FOUND THEN
+            INSERT INTO task_runs (
+                stream_id, task_name, consumer_group, consumer_name, status, result, error_message, 
+                start_time, end_time, worker_id, duration, execution_time,
+                created_at, updated_at
+            ) VALUES (
+                (v_record->>'stream_id')::TEXT,
+                (v_record->>'task_name')::TEXT,
+                (v_record->>'consumer_group')::TEXT,
+                (v_record->>'consumer_name')::TEXT,
+                COALESCE((v_record->>'status')::TEXT, 'pending'),
+                (v_record->>'result')::jsonb,
+                (v_record->>'error_message')::TEXT,
+                (v_record->>'started_at')::TIMESTAMPTZ,
+                (v_record->>'completed_at')::TIMESTAMPTZ,
+                (v_record->>'worker_id')::TEXT,
+                (v_record->>'duration')::DOUBLE PRECISION,
+                (v_record->>'execution_time')::DOUBLE PRECISION,
+                CURRENT_TIMESTAMP,
+                CURRENT_TIMESTAMP
+            );
+        END IF;
+        
+        v_count := v_count + 1;
+    END LOOP;
+    
+    RETURN v_count;
+END;
+$$ LANGUAGE plpgsql;
+
+-- 批量插入tasks的优化函数
+CREATE OR REPLACE FUNCTION batch_insert_tasks(
+    p_records jsonb
+) RETURNS INTEGER AS $$
+DECLARE
+    v_inserted INTEGER;
+BEGIN
+    -- 使用单个INSERT语句批量插入，忽略冲突
+    WITH data AS (
+        SELECT 
+            (value->>'stream_id')::TEXT as stream_id,
+            (value->>'queue')::TEXT as queue,
+            (value->>'namespace')::TEXT as namespace,
+            (value->>'scheduled_task_id')::TEXT as scheduled_task_id,
+            (value->>'payload')::jsonb as payload,
+            (value->>'priority')::INTEGER as priority,
+            (value->>'created_at')::TIMESTAMPTZ as created_at,
+            (value->>'source')::TEXT as source,
+            (value->>'metadata')::jsonb as metadata
+        FROM jsonb_array_elements(p_records)
+    )
+    INSERT INTO tasks (stream_id, queue, namespace, scheduled_task_id, payload, priority, created_at, source, metadata)
+    SELECT * FROM data
+    ON CONFLICT DO NOTHING;
+    
+    GET DIAGNOSTICS v_inserted = ROW_COUNT;
+    RETURN v_inserted;
+END;
+$$ LANGUAGE plpgsql;
+
+-- 批量清理已完成任务的函数
+CREATE OR REPLACE FUNCTION cleanup_completed_tasks(
+    p_stream_ids TEXT[]
+) RETURNS INTEGER AS $$
+DECLARE
+    v_deleted INTEGER;
+BEGIN
+    -- 批量删除已完成的任务
+    DELETE FROM task_runs 
+    WHERE stream_id = ANY(p_stream_ids)
+      AND status IN ('success', 'failed', 'timeout', 'skipped', 'cancelled');
+    
+    GET DIAGNOSTICS v_deleted = ROW_COUNT;
+    RETURN v_deleted;
+END;
+$$ LANGUAGE plpgsql;
+
+-- ========================================
+-- 13. 创建额外的优化索引
+-- ========================================
+
+-- 为task_runs的UPSERT操作创建覆盖索引（如果不存在）
+CREATE INDEX IF NOT EXISTS idx_task_runs_upsert 
+ON task_runs (stream_id, consumer_group) 
+INCLUDE (status, updated_at);
+
+-- ========================================
+-- 14. 创建默认数据
+-- ========================================
+-- 这里可以插入一些默认的配置或测试数据
+
+-- ========================================
+-- 完成
+-- ========================================
+-- 注意：最后的 RAISE NOTICE 可能在某些环境中不支持，可以注释掉
+-- RAISE NOTICE 'Database initialization completed successfully!';