At present, `load_to_singlt_tablet` import implementation refers to simple random number remainder, which cannot achieve true averaging. This will lead to uneven disk IO and uneven use of cluster resources. To solve this problem, we are preparing to implement round-robin for each partition tablet imported each time, in order to achieve average load to each tablet.
When generating the load query plan, the tablet index record currently imported is passed to BE.
Add a deamon task in FE to regularly clean up the `loadTabletRecordMap`. The map will get the bucket_number of the partition and update the `load_tablet_index` when `getCurrentLoadTabletIndex`.
Disable infer expr column name when query on old optimizer.
This bug is be brought in #24990
if your query SQL is
select id, name, sum(target) FROM db_test.table_test2 group by id, name;
the result column name when query is as following:
|id|name |sum(cast(target as DOUBLE))|
when you create view as following:
CREATE VIEW v1 as select id, name, sum(target) FROM db_test.table_test2 group by id, name;
then query the view v1, the result is as following:
|id|name |__sum_2|
Problem:
when running pipeline, we get randomly failed of test_leading
Reason:
physical distribute was generated and choosed to be the best plan because we can not get any statistic information of empty table. So we would get some unexpect result because we can not expect the order in memo
Solved:
Add statistic of columns used in test_leading, try repeatly in pipeline
this pr
1. fix use podarray push_back() with back() will make heap_use_after_free when podarray is reach capacity which would may make heap free
2. add cases for csv format for nested types. and csv file has two define which are without quote or just like json text
The user has configured the parameter lower_case_table_names, which ignores the case of the table name. When executed on the SQL client, the table name can be queried in both case.
But when using Connector to read doris data, the table names must be in the same case, otherwise an error will be reported.
[fix](case) change dynamic_partition.time_unit from day to month to avoid the error that the intert data not in partition
Co-authored-by: stephen <hello-stephen@qq.com>
By utilizing the zonemap index of the segment, we can ascertain if a predicate is always true. For example, if the segment’s maximum value is 100 and the predicate is col < 101, then this predicate is always true for this segment.
