[Bug] window assigns error
- 主要语言
- Java
- 星标
- 6.4k
- 派生
- 1.2k
- 平均合并
- 1 天 23 小时
- 30 天内合并 PR
- 115
描述
### Search before asking
- [X] I searched in the [issues](https://github.com/apache/iotdb/issues) and found nothing similar.
### Version

操作系统:CentOS 7.9
IoTDB:1.1.1
### Describe the bug and provide the minimal reproduce step
window assigns in group by is different in same logic
### execute insert sql
`insert into root.hess.test.AK1407120700022.testdata (timestamp, epv_total, eload_total) VALUES (2592000000,1,1),(5011199999,2,2),(5011200000,3,3),(7689599999,4,4),(7689600000,5,5)`
`insert into root.hess.test.AK1407120700022.testdata (timestamp, epv_total, eload_total) VALUES (2563200000,10,10),(4982399999,20,20),(4982400000,30,30),(7660799999,40,40),(7660800000,50,50)`
### then execute query sql
`select SUM(epv_total) as pv,SUM(eload_total) as load from root.hess.**.AK1407120700022.** group by ([1970-02-28T08:00:00, 1970-05-31), 1mo , 2mo)`
`select SUM(epv_total) as pv,SUM(eload_total) as load from root.hess.**.AK1407120700022.** group by ([1970-02-28T08:00:00, 1970-05-31), 1mo , 1mo)`
### What did you expect to see?
IoTDB> select SUM(epv_total) as pv,SUM(eload_total) as load from root.hess.**.AK1407120700022.** group by ([1970-02-28T08:00:00, 1970-05-31), 1mo , 2mo)
+-----------------------------+----+----+
| Time| pv|load|
+-----------------------------+----+----+
|1970-02-28T08:00:00.000+08:00|97.0|97.0|
|1970-04-30T08:00:00.000+08:00|null|null|
+-----------------------------+----+----+
Total line number = 2
It costs 0.025s
IoTDB> select SUM(epv_total) as pv,SUM(eload_total) as load from root.hess.**.AK1407120700022.** group by ([1970-02-28T08:00:00, 1970-05-31), 1mo , 1mo)
+-----------------------------+----+----+
| Time| pv|load|
+-----------------------------+----+----+
|1970-02-28T08:00:00.000+08:00|97.0|97.0|
|1970-03-31T08:00:00.000+08:00| 5.0| 5.0|
|1970-04-30T08:00:00.000+08:00|null|null|
+-----------------------------+----+----+
Total line number = 3
### What did you see instead?

### Anything else?
_No response_
### Are you willing to submit a PR?
- [X] I'm willing to submit a PR!
贡献指南
调研方向
在插入列出的行后,针对 IoTDB 1.1.1 运行提供的两个 GROUP BY 查询,并比较 1mo, 2mo 和 1mo, 1mo 的结果。追踪查询引擎中的窗口分配行为;完成标准是等效的窗口逻辑产生一致的区间分配,并且复现查询返回预期的行。
由索引模型根据 Issue 内容生成。
评估
- 技术栈
- java, sql
- 领域
- databases
- Issue 类型
- 缺陷
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 活跃度
- 停滞
- 描述清晰度
- 基本清楚
- 新手友好度
- 35/100