ClickHouse and his friends (3) MySQL Protocol and Write call stack

Original source: https://bohutang.me/2020/06/08/clickhouse-and-friends-mysql-protocol-write-stack/

The MySQL Protocol and Read call in the previous article introduced the call stack of a ClickHouse query statement. This article continues to introduce the call stack written and opened the whole.

Write request

  1. Build a table:

mysql> CREATE TABLE test(a UInt8, b UInt8, c UInt8) ENGINE=MergeTree() PARTITION BY (a, b) ORDER BY c;
Query OK, 0 rows affected (0.03 sec)
  1. data input:

INSERT INTO test VALUES(1,1,1), (2,2,2);

Call stack analysis

1. Get the storage engine OutputStream

DB::StorageMergeTree::write(std::__1::shared_ptr<DB::IAST> const&, DB::Context const&) StorageMergeTree.cpp:174
DB::PushingToViewsBlockOutputStream::PushingToViewsBlockOutputStream(std::__1::shared_ptr<DB::IStorage> const&, DB::Context const&, std::__1::shared_ptr<DB::IAST> const&, bool) PushingToViewsBlockOutputStream.cpp:110
DB::InterpreterInsertQuery::execute() InterpreterInsertQuery.cpp:229
DB::executeQueryImpl(const char *, const char *, DB::Context &, bool, DB::QueryProcessingStage::Enum, bool, DB::ReadBuffer *) executeQuery.cpp:364
DB::executeQuery(DB::ReadBuffer&, DB::WriteBuffer&, bool, DB::Context&, std::__1::function<void (std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)>) executeQuery.cpp:696
DB::MySQLHandler::comQuery(DB::ReadBuffer&) MySQLHandler.cpp:311
DB::MySQLHandler::run() MySQLHandler.cpp:141

2. Assemble InputStream from SQL

(1,1,1), (2,2,2) How to assemble it into an inputstream structure?

DB::InputStreamFromASTInsertQuery::InputStreamFromASTInsertQuery(std::__1::shared_ptr<DB::IAST> const&, DB::ReadBuffer*, 
DB::InterpreterInsertQuery::execute() InterpreterInsertQuery.cpp:300
DB::executeQueryImpl(char const*, char const*, DB::Context&, bool, DB::QueryProcessingStage::Enum, bool, DB::ReadBuffer*) executeQuery.cpp:386
DB::MySQLHandler::comQuery(DB::ReadBuffer&) MySQLHandler.cpp:313
DB::MySQLHandler::run() MySQLHandler.cpp:150

then

res.in = std::make_shared<InputStreamFromASTInsertQuery>(query_ptr, nullptr, query_sample_block, context, nullptr);
res.in = std::make_shared<NullAndDoCopyBlockInputStream>(res.in, out_streams.at(0));

Construct a Block through the copyData method of NullAndDoCopyBlockInputStream:

DB::ValuesBlockInputFormat::readRow(std::__1::vector<COW<DB::IColumn>::mutable_ptr<DB::IColumn>, std::__1::allocator<COW<DB::IColumn>::mutable_ptr<DB::IColumn> > >&, unsigned long) ValuesBlockInputFormat.cpp:93
DB::ValuesBlockInputFormat::generate() ValuesBlockInputFormat.cpp:55
DB::ISource::work() ISource.cpp:48
DB::InputStreamFromInputFormat::readImpl() InputStreamFromInputFormat.h:48
DB::IBlockInputStream::read() IBlockInputStream.cpp:57
DB::InputStreamFromASTInsertQuery::readImpl() InputStreamFromASTInsertQuery.h:31
DB::IBlockInputStream::read() IBlockInputStream.cpp:57
void DB::copyDataImpl<DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::atomic<bool>*)::$_0&, void (&)(DB::Block const&)>(DB::IBlockInputStream&, DB::IBlockOutputStream&, DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::atomic<bool>*)::$_0&, void (&)(DB::Block const&)) copyData.cpp:26
DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::atomic<bool>*) copyData.cpp:62
DB::NullAndDoCopyBlockInputStream::readImpl() NullAndDoCopyBlockInputStream.h:47
DB::IBlockInputStream::read() IBlockInputStream.cpp:57
void DB::copyDataImpl<std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&>(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&) copyData.cpp:26
DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&) copyData.cpp:73
DB::executeQuery(DB::ReadBuffer&, DB::WriteBuffer&, bool, DB::Context&, std::__1::function<void (std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)>) executeQuery.cpp:785
DB::MySQLHandler::comQuery(DB::ReadBuffer&) MySQLHandler.cpp:313
DB::MySQLHandler::run() MySQLHandler.cpp:150

3. Assemble the OutputStream

DB::InterpreterInsertQuery::execute() InterpreterInsertQuery.cpp:107
DB::executeQueryImpl(const char *, const char *, DB::Context &, bool, DB::QueryProcessingStage::Enum, bool, DB::ReadBuffer *) executeQuery.cpp:364
DB::executeQuery(DB::ReadBuffer&, DB::WriteBuffer&, bool, DB::Context&, std::__1::function<void (std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)>) executeQuery.cpp:696
DB::MySQLHandler::comQuery(DB::ReadBuffer&) MySQLHandler.cpp:311
DB::MySQLHandler::run() MySQLHandler.cpp:141

Assembly sequence:

  1. NullAndDoCopyBlockInputStream

  2. CountingBlockOutputStream

  3. AddingDefaultBlockOutputStream

  4. SquashingBlockOutputStream

  5. PushingToViewsBlockOutputStream

  6. MergeTreeBlockOutputStream

4. Write to OutputStream

DB::MergeTreeBlockOutputStream::write(DB::Block const&) MergeTreeBlockOutputStream.cpp:17
DB::PushingToViewsBlockOutputStream::write(DB::Block const&) PushingToViewsBlockOutputStream.cpp:145
DB::SquashingBlockOutputStream::finalize() SquashingBlockOutputStream.cpp:30
DB::SquashingBlockOutputStream::writeSuffix() SquashingBlockOutputStream.cpp:50
DB::AddingDefaultBlockOutputStream::writeSuffix() AddingDefaultBlockOutputStream.cpp:25
DB::CountingBlockOutputStream::writeSuffix() CountingBlockOutputStream.h:37
DB::copyDataImpl<DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::atomic<bool>*)::<lambda()>&, void (&)(const DB::Block&)>(DB::IBlockInputStream &, DB::IBlockOutputStream &, <lambda()> &, void (&)(const DB::Block &)) copyData.cpp:52
DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::atomic<bool>*) copyData.cpp:138
DB::NullAndDoCopyBlockInputStream::readImpl() NullAndDoCopyBlockInputStream.h:57
DB::IBlockInputStream::read() IBlockInputStream.cpp:60
void DB::copyDataImpl<std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&>(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&) copyData.cpp:29
DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&) copyData.cpp:154
DB::executeQuery(DB::ReadBuffer&, DB::WriteBuffer&, bool, DB::Context&, std::__1::function<void (std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)>) executeQuery.cpp:748
DB::MySQLHandler::comQuery(DB::ReadBuffer&) MySQLHandler.cpp:311
DB::MySQLHandler::run() MySQLHandler.cpp:141

Through the copyData method, the data is transparently transmitted between the OutputStream layers, all the way to the MergeTreeBlockOutputStream.

5. Return to Client

DB::MySQLOutputFormat::finalize() MySQLOutputFormat.cpp:62
DB::IOutputFormat::doWriteSuffix() IOutputFormat.h:78
DB::OutputStreamToOutputFormat::writeSuffix() OutputStreamToOutputFormat.cpp:18
DB::MaterializingBlockOutputStream::writeSuffix() MaterializingBlockOutputStream.h:22
void DB::copyDataImpl<std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&>(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&) copyData.cpp:52
DB::copyData(DB::IBlockInputStream&, DB::IBlockOutputStream&, std::__1::function<bool ()> const&, std::__1::function<void (DB::Block const&)> const&) copyData.cpp:154
DB::executeQuery(DB::ReadBuffer&, DB::WriteBuffer&, bool, DB::Context&, std::__1::function<void (std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)>) executeQuery.cpp:748
DB::MySQLHandler::comQuery(DB::ReadBuffer&) MySQLHandler.cpp:311
DB::MySQLHandler::run() MySQLHandler.cpp:141

to sum up

INSERT INTO test VALUES(1,1,1), (2,2,2);

First, the kernel parses the SQL statement to generate an AST, and obtains the Interpreter: InterpreterInsertQuery according to the AST. Secondly, Interpreter adds the corresponding OutputStream in turn. Then read data from InputStream and write it to OutputStream. Stream will penetrate layer by layer and write to the underlying storage engine. Finally, write to Socket Output and return the result.

ClickHouse's OutputStream layout is still more complicated, lacking pipeline-like scheduling and orchestration, but due to the solidification of the model, it seems clear at present.

In-text link

The full text is over.

Enjoy ClickHouse :)

Teacher Ye's "MySQL Core Optimization" class has been upgraded to MySQL 8.0, scan the code to start the journey of MySQL 8.0 practice

Guess you like

Origin blog.csdn.net/n88Lpo/article/details/108934551