There are several configs related to tcmalloc, users do know how to config them. Actually users just want two modes, performance or compact, in performance mode, users want doris run query and load quickly while in compact mode, users want doris run with less memory usage.
If we want to config tcmalloc individually, we can use env variables which are supported by tcmalloc.
When upgrade 1.2 version from 1.1, FE version will don't match BE version for a period of time. After upgrade BE and doing schema change, BE will use a field desc_tbl that add in 1.2 version FE. BE will coredump because the field desc_tbl is nullptr. So it need to refuse the request.
- this pr is used to fix the be core dump when import array.
- before the change, we import array by rapidjson string will core dump under the non-vectorized scenario.
- after the change, we can import array by rapidjson string successfully.
when enable_nereids_planner=false and enable_fallback_to_origin=false, FE throws exception for all select statement.
Expected: when enable_nereids_planner=false, all valid query execution success
1.
In the previous implementation, the max thread num of olap scanner was set relatively small, such as 3.
which would slow down some of queries.
In this PR, I changed the max thread num to a quarter of the scaner thread pool(default is 12),
which is less than the old scan node's max thread num, but larger than the previous implementation.
The upper limit of the max thread num of the old scan node is too high, which is not reasonable.
2.
Lower down the number of pre allocated free blocks.
1.remove quick_compaction's rowset pick policy, call cu compaction when trigger
quick compaction
2. skip tablet's compaction task when compaction score is too small
Co-authored-by: yixiutt <yixiu@selectdb.com>
1.
In the "explain" result of SQL, the table name in `ScanNode` should be full qualified with dbname.
And for olap scan node, the selected index name should not be "null".
2.
Remove `tpch_sf1_p1/tpch_sf1/nereids/` in regression test, it will be fixed later.
PR(https://github.com/apache/doris/pull/13404) introduced that ParquetReader
will break up batch insertion when encountering null values, which leads to the bad performance
compared to OrcReader.
So this PR has pushed null map into decode function, reduce the time of virtual function call
when encountering null values.
Further more, reuse hdfsFS among file readers to reduce the time of building connection to hdfs.
