ZHANG Hao

@tqchen @tmoreau89 Could you kindly help review this PR? Thanks.

> Thanks @zhanghaohit, for catching this error. I agree with the fact that the dependence should not be hardcoded. However in order to not to add too many parameters in...

> Perhaps one way to look at this is to start with DISTANCE=3 by default. And if the II>1 for the GEMM, we issue a warning telling the user to...

rebase or merge from `main` branch

The problem still exists. See [select * from __INTERNAL_DB.PRE_AGG_META_INFO](https://github.com/4paradigm/OpenMLDB/issues/2464#issuecomment-1254596836) in #2464

> 版本：0.6.2 集群1060份请求样本数据压测，去除10d窗口qps 4w，带10d窗口性能降低至30qps，长窗口优化是否是处理这方面 数据内容为20220814后一周约800w条数据（数据分布整体均匀），超时时间设定为60d @uttinie 需要几点帮忙确认的地方： - 先执行deploy with long_windows options，再导入数据? - 连接sql_client查看预聚合表信息`select * from __INTERNAL_DB.PRE_AGG_META_INFO;` - 连接ns_client， ``` use __PRE_AGG_DB showtable ```

> pre_demo_db_demo9option_ab_w10d_count_where_device_name_emui_ver @uttinie - 导入数据是否按照时间顺序，由小到大导入的？ - `pro_ver`的取值有多少？只有`12.0`和`6.1`两个取值么？ - 可以在ns client上，执行 ``` use demo_db showtable ``` - 在sql client上执行 ``` select * from __PRE_AGG_DB.pre_demo_db_demo9option_ab_w10d_count_where_device_name_emui_ver limit 100; select * from __PRE_AGG_DB.pre_demo_db_demo9option_cat_w10d_count_where_app_ver_emui_ver limit...

> https://github.com/4paradigm/OpenMLDB/blob/309dc04cdc16f5529ea09122c4e0d2849d1df90a/src/storage/aggregator.cc#L359-L363 > > There seems to be nothing to do, you can remove the warning log. here if we found a `kDelete` entry, we shall do `Delete(key)` again, right?...

merged after 060 release

check if there is ut to check: after gc, the data is actually deleted