[STORAGE]
1 fix mem fix mem leak when calling string builder.get_dictionary_page;
2 fix delete invalid mem addr in bitshuffleBuilder when no array grow happends
when bitshuffleBuilder didn't grow array, the data page which not use new to allocate will be
returned to ColumnWriter.
When ColumnWriter destructs, the data page will be deleted,this causes core dump
[Storage][SegmentV2]
Currently `segment_v2::Segment::open` will eagerly initialize all column readers, regardless of whether the column is queried or not. Initializing `segment_v2::ColumnReader` incurs additional I/O cost to read ordinal index and zonemap index and should be delayed to the time it's needed.
[Storage]
This PR implements thread-safe `Rowset::load()` for both AlphaRowset and BetaRowset. The main changes are
1. Introduce `DorisCallOnce<ReturnType>` to be the replacement for `DorisInitOnce` . It works for both Status and OLAPStatus.
2. `segment_v2::ColumnReader::init()` is now implemented by DorisCallOnce.
3. `segment_v2::Segment` is now created by a factory open() method. This guarantees all Segment instances are in opened state.
4. `segment_v2::Segment::_load_index()` is now implemented by DorisCallOnce.
5. Implement thread-safe load() for AlphaRowset and BetaRowset
After replacing Arena with MemPool, we can achieve one copy for string
value read from segment v2. We can exchange MemPool's chunk between
RowBlockV2 and RowBlock. This change only replace Arena, this work will
be done in other change list.
[Bug][BetaRowset] fix beta rowset read slowly with limit
beta rowset do not update raw_rows_read in statistics and will read all
data in tablet when query with limit, which lead to long query time.
Now Env has unify all environment operation, such as file operation.
However some of our old functions don't leverage it. This change unify
FileUtils::scan_dir to use Env's function.
In Storage Engine GC, TabletManger use std::regex to extract tablet id
and schema hash from path. But it will construct regex pattern for
every path to check, this is a huge waste. This change list make this
pattern a global static pattern, and replace it with RE2, which has
better performance.
Leverage gitattributes to enable auto convert end-of-line to LF when
checking in. Convert already exist CRLF to LF by removing all files and
checking out with new .gitattributes file. Except .gitattributes, all
files are only modified at the end of line.
major change
change data format of binary dict page, appending (dict page data) and (dict page offset) to binary dict page;
add new decoding method for new binary dict page format
add ut for segment test
set the elements of initial array to 0 ,when calling arena.AllocateNewBlock
hard code way to choose dict coding for string
0919 commit major change
change dict file format:when saving binary dict page, separate dict page from dict page,one dict page may have multi data pages;when reading a binary dict page,one ColumnReader keeps one dict page
loading dict when calling column_reader._read_page
3.rollback BinaryDictPage
no longer using memset(0) to inital column_zonemap.max_value
0926 17 commit major change
init column_zone_map min value column_zone_map slice's data array;
set char/varchar column_zone_map'max value size to 0
add ut for char column zone map query hit/miss
0929 10 commit major change
allocate mem for column_zone_map 's max and min value
direct copy content to column_zone_map's max and min value
Remove the default constructor for UniqueID
Add a gen_uid method in UniqueId. If need to generate a new uid, users should call this api explicitly.
Reuse boost random generator not generate a new one every time.
In this change list
1. validate HLL column when loading data, if data is invalid, this row
will be filtered.
2. seems as empty HLL when serializing invalid type of HLL data, with
this change, all ingested data will be valid.
3. seems as empty HLL when deserializing nullptr or invalid type of HLL data.
With this change, dirty data can be handled normally.
4. rename function empty_hll to hll_empty.
5. disable memtable_flush_execute_test because this will fails
sometimes. When tearing down, some thread is not joined, and they will
visit destroyed resource, which is invalid.
The current load process is:
Tablet Sink -> Tablet Channel Mgr -> Tablets Channel -> Delta Writer -> MemTable -> Flush to disk
In the path of Tablets Channel -> DeltaWriter -> MemTable -> Flush to disk, the following operations are performed:
Insert tuple into different memtables according to tablet ID
When the memtable size reaches the threshold, it is written to disk.
The above operations are equivalent to single thread execution for a single load task.
In fact, the insertion of memtable and the flush of memtable can be executed synchronously.
Perform these operation in single thread prevents the insertion of memtable from being delayed due to slow disk writing.
In the new implementation, I added a MemTableFlushExecutor class with a set of flush queues and corresponding worker threads.
By default, each data directory uses two worker threads for flush, which can be modified by the parameter flush_thread_num_per_store of BE.
DeltaWriter will push the full memtable to MemTableFlushExecutor for flush operation and generate a new memtable for receiving new data.
This design can improve the performance of load large files.
In single host testing, the time to load a 1GB text file is reduced from 48 seconds to 29 seconds.
Now size of HyperLogLog struct is so large that it lead the rowset is
too small when ingesting data. In this CL, registers in HyperLogLog are
only created when it is needed. When ingesting data, it's normal case
that there are only few values in one HyperLogLog.
In Doris, one block has 1024 rows.
1. If the previous ScanKey scan rows multiple blocks,
and also the final block has 1024 rows just right.
2. The current ScanKey scan rows with number less than one block.
Under the two conditions, if not seek block, the position of prefix shortkey columns is wrong.
I add ChunkAllocator in this CL to put unused memory chunk to a chunk
pool other than return it to system allocator. Now we only change
MemPool's chunk allocation and free to this.
And two configuration are introduduced too. 'chunk_reserved_bytes_limit'
is the limit of how many bytes this chunk pool can reserve in total and
its default value is 2147483648(2GB). 'use_mmap_allocate_chunk': if
chunk is allocated via mmap and default value is false.
And in my test case with default configuration a simple like
"select * from table limit 10", this can improve throughput from 280 QPS
to to 650 QPS. And when I config 'chunk_reserved_bytes_limit' to 0,
which means this is disabled, the throughput is the same with origin's.
