feat: optimize connect_process_and_networks

[taloric]

• optimze connect_process_and_networks func cost
  1. use related_flow_index_map to optimize O(n^2) loop scan
  2. use single _id IN(x) instead of _id IN(xx) and _id IN(xx) to optimize query performance
  3. use bit calculation instead of string array for related_types

---

• 优化 connect_process_and_networks 函数
  1. 使用related_flow_index_map 来优化原有的 O(n^2) 扫描，减少扫描量
  2. 使用一个单一的 _id in (xx) 而不是多个 _id in (xx) and _id in (xx) 来提升查询效果
  3. 计算 related_types 时，使用位运算取代原有的字符串数组实现

[taloric]

测试结果：

1. 测试集：8000 个 app_span，会发生 process_roots 与 process_leafs 双重连接，修改前消耗：

   
   real_time: 2024-09-11 10:05:25
   start_time: 2024-09-11 10:05:00, end_time: 2024-09-11 10:23:00
   cost: 457.8415722846985 s
   len(result['DATA']['tracing']): 8000

connect_process_and_networks time: 369.36000752449036

修改后：
```text
real_time: 2024-09-11 10:05:25
start_time: 2024-09-11 10:05:00, end_time: 2024-09-11 10:23:00
cost: 13.16390872001648 s
len(result['DATA']['tracing']): 8000

connect_process_and_networks time: 2.6919126510620117

2. 测试集：spring-boot-demo，82个 span，会发生多次连接

修改前:

real_time: 2024-09-11 13:11:50
start_time: 2024-09-11 13:09:50, end_time: 2024-09-11 13:13:50
cost: 10.657386779785156 s
len(result['DATA']['tracing']): 82

connect_process_and_networks time: 0.13371014595031738

修改后：

real_time: 2024-09-11 13:11:50
start_time: 2024-09-11 13:09:50, end_time: 2024-09-11 13:13:50
cost: 8.954121589660645 s
len(result['DATA']['tracing']): 82

connect process and network cost time: 0.0021207332611083984

3. 测试集，grpc-demo，111 个 span 修改前：

real_time: 2024-09-11 14:34:40
start_time: 2024-09-11 14:32:40, end_time: 2024-09-11 14:36:40
cost: 3.981588363647461 s
len(result['DATA']['tracing']): 111

connect_process_and_networks cost time: 0.02313542366027832

修改后：

real_time: 2024-09-11 14:34:40
start_time: 2024-09-11 14:32:40, end_time: 2024-09-11 14:36:40
cost: 3.298126220703125 s
len(result['DATA']['tracing']): 111

connect process and network cost time: 0.0023322105407714844

[taloric]

总结：对于大数据集，提升明显，对于小数据集，也略有收益

数据集	测试对象	修改前	修改后	总体提升
8000 span	整体耗时	457.8415722846985 s	13.16390872001648 s	97%
82 span	整体耗时	10.657386779785156 s	8.954121589660645 s	15%
111 span	整体耗时	3.981588363647461 s	3.298126220703125 s	17%
8000 span	connect_process_and_networks 耗时	369.36000752449036	2.6919126510620117	99%
82 span	connect_process_and_networks 耗时	0.13371014595031738	0.0021207332611083984	98%
111 span	connect_process_and_networks 耗时	0.02313542366027832	0.0023322105407714844	89%