1. Spark can set the timestamp precision by the following configuration:
spark.sql.parquet.outputTimestampType = INT96(NANOS), TIMESTAMP_MICROS, TIMESTAMP_MILLIS
DATETIME V1 only keeps the second precision, DATETIME V2 keeps the microsecond precision.
2. If using DECIMAL V2, the BE saves the value as decimal128, and keeps the precision of decimal as (precision=27, scale=9). DECIMAL V3 can maintain the right precision of decimal
Stream load will ignore invisible columns if no http header columns
specified, but in some case user cannot get all columns if columns
changed frequently。
Add a hidden_columns header to support hidden columns import。User can
set hidden_columns such as __DORIS_DELETE_SIGN__ and add this column
in stream load data so we can delete this line.
For example:
curl -u root -v --location-trusted -H "hidden_columns: __DORIS_DELETE_SIGN__" -H
"format: json" -H "strip_outer_array: true" -H "jsonpaths: [\"$.id\",
\"$.name\",\"$.__DORIS_DELETE_SIGN__\"]" -T 1.json
http://{beip}:{be_port}/api/test/test1/_stream_load
Co-authored-by: yixiutt <yixiu@selectdb.com>
when config::enable_simdjson_parser=true in vec streamload, may lead to core dump when json input invalid format string like '{ "a', or all the fields is null like '{}', this may lead to simdjson lib throw some unhandled expection like `Objects and arrays can only be iterated when they are first encountered`.We should take care of these cases
Signed-off-by: eldenmoon <15605149486@163.com>
The column id check in VSlotRef::execute function before is too strict for fuzzy test to continuously produce random query. Temporarily loosen the check logic.
Moreover, there exists some careless call to VExpr::get_const_col, it might return a nullptr but not every function call checks if it's valid. It's an underlying problem.
some feature:
1. add min max key in segment footer to speed up get_row_ranges_by_keys
2. do not load pk bloom filter in query
Co-authored-by: yixiutt <yixiu@selectdb.com>
In engine_clone_task.cpp, it use tablet->tablet_schema() to create rowset, but in the method, it need a lock that already locked in engine_clone_task.cpp:514. It use cloned_tablet_meta->tablet_schema() originally, but modified in #11131. It need to revert to use cloned_tablet_meta->tablet_schema().
Currently we use rapidjson to parse json document, It's fast but not fast enough compare to simdjson.And I found that the simdjson has a parsing front-end called simdjson::ondemand which will parse json when accessing fields and could strip the field token from the original document, using this feature we could reduce the cost of string copy(eg. we convert everthing to a string literal in _write_data_to_column by sprintf, I saw a hotspot from the flamegrame in this function, using simdjson::to_json_string will strip the token(a string piece) which is std::string_view and this is exactly we need).And second in _set_column_value we could iterate through the json document by for (auto field: object_val) {xxx}, this is much faster than looking up a field by it's field name like objectValue.FindMember("k1").The third optimization is the at_pointer interface simdjson provided, this could directly get the json field from original document.
Disable Chunk Allocator in Vectorized Allocator, this will reduce memory cache.
For high concurrent queries, using Chunk Allocator with vectorized Allocator can reduce the impact of gperftools tcmalloc central lock.
Jemalloc or google tcmalloc have core cache, Chunk Allocator may no longer be needed after replacing gperftools tcmalloc.
The bug is caused by use _num_rows_read for limit check. _num_rows_read is count of rows read from storage, but may be filtered by filter_block for WHERE predicate.
Add a _num_rows_return, which is rows after filter_block for WHERE predicate, for count for really returned rows.
The memory value automatically tracked by the tcmalloc hook in the DeltaWriter is smaller than the value recorded manually in the memtable, because the first 4096-byte Chunk requested by each MemPool when the memtable is initialized is not tracked to the DeltaWriter by the hook.
The values of the two are not equal, causing the mem_consumption() == _mem_table->memory_usage branch judgment to fail.
Currently, only the virtual memory used by the query can be tracked through the tcmalloc hook. When the memory is not fully used after the application, the recorded virtual memory will be larger than the physical memory.
At present, it is mainly because PODArray does not memset 0 when applying for memory, and blocks applied for through PODArray in places such as VOlapScanNode::_free_blocks are usually used for memory reuse and cannot be fully used.
