When performing storage_medium_migration, the header may be already droped.
In this scenario, header returned will be null pointer.
Saving data to a null pointer will cause core dump.
* This commit has brought contribution to streaming mini load
The operation of streaming mini load is sames as previous. Also, user can check the load by frontend.
The difference is that streaming mini load finish the task before reply of REST API while the non-streaming only register a load.
* When updating doris
Updating fe or be firstly are also supported. After fe and be are updated, the streaming mini load will take effect.
* For multi mini load
The non-streaming mini load still has been used by multi mini load. The behavior of multi mini load has not been changed.
* Add a interface named isSupportedFunction
This function is used to protect the correctness of new feature which consists of be and fe during updaing.
In streaming load, one version will generate multi SegmentGroups.
Upon creating rollup, the previous code only check version exists or not.
Instead, every SegmentGroup should be checked independently.
This change adds a load property named strict_mode which is used to prohibit the incorrect data.
When it is set to false, the incorrect data will be loaded by NULL just like before.
When it is set to true, the incorrect data which belongs to a column without expr will be filtered.
The strict_mode is supported in broker load v2 now. It will be supported in stream load later.
Before RowBlock will reserve memory for all columns in schema, even if
it is not queried. Which will cause bad performance when quering wide
table.
In this patch, RowBlock will reserve memory for needed columns. In a
case, this reduce ConvertBatchTime from 10s to 60ms when quering a wide
table who has 178 columns.
#1236
When I declared that the compilation mode was BUILD_TYPE=LSAN, there was a memory leak after running doris.
be.out:
Direct leak of 32816 byte(s) in 1 object(s) allocated from:
#0 0x1089666 in __interceptor_malloc ../../../../libsanitizer/lsan/lsan_interceptors.cc:53
#1 0x7ff459547280 in __alloc_dir (/lib64/libc.so.6+0xc0280)
SUMMARY: LeakSanitizer: 32816 byte(s) leaked in 1 allocation(s).
* Fix the bug of += decimal in olap engine
[ISSUE-1225] This change fix the olap engine bug of decimal agg. Using ^ instead of * to judge result is less then zero.
The result of * will be less then zero when the result is overflow. So the answer of += is incorrect.
The non-streaming hint of insert into will use the streamin plan which is same as the plan of stream insert.
It will also record the load info and return the label of insert stmt.
The partition is supportted in insert into stmt. The result which meet the target partitions will be loaded.
The introduction of example has been changed especially non-streaming insert.
Also, the param of partition_names is added in sql syntax which is used to declare the target partition_names in target table.
Change META_VERSION to 50
When creating new tablet by alter tablet task,
next_unique_id will increase on the base old tablet.
If next_unique_id is eqaul to zero, it will cause that
ColumnDataMessage not match with tablet meta.
1. get_json_xxx() now support using quoto to escape dot
2. Implement json_path_prepare() function to preprocess json_path
Performance of get_json_string() on 1000000 rows reduces from 2.27s to 0.27s
