thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-21 15:12:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

better processing stage

Browse Source
This commit is contained in:
Nikita Mikhaylov 2021-04-07 17:43:34 +03:00
parent 587fbdd10d
commit 2555ae5d3f
3 changed files with 36 additions and 45 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/Client/Connection.h
View File

@ -65,6 +65,8 @@ struct Packet
Progress progress;
BlockStreamProfileInfo profile_info;
std::vector<UUID> part_uuids;
/// String describes an identifier for a request.
/// Used for dynamic distributed data processing (S3 downloading)
String read_task_request;
Packet() : type(Protocol::Server::Hello) {}

74
src/Storages/StorageS3Distributed.cpp
View File

@ -90,8 +90,8 @@ public:
const StorageS3SourceBuilder & s3_builder_)
: SourceWithProgress(getHeader(s3_builder_.sample_block, s3_builder_.need_path, s3_builder_.need_file))
, initial_query_id(initial_query_id_)
, s3_source_builder(s3_builder_)
, cli_builder(client_auth_builder_)
, s3b(s3_builder_)
, cab(client_auth_builder_)
, read_task_callback(read_task_callback_)
{
createOrUpdateInnerSource();
@ -128,7 +128,7 @@ private:
}
catch (...)
{
tryLogCurrentException(&Poco::Logger::get("StorageS3SequentialSource"));
tryLogCurrentException(&Poco::Logger::get(getName()));
throw;
}
}
@ -143,21 +143,15 @@ private:
auto client_auth = StorageS3::ClientAuthentificaiton{
next_uri,
cli_builder.access_key_id,
cli_builder.secret_access_key,
cli_builder.max_connections,
cab.access_key_id,
cab.secret_access_key,
cab.max_connections,
{}, {}};
StorageS3::updateClientAndAuthSettings(s3_source_builder.context, client_auth);
StorageS3::updateClientAndAuthSettings(s3b.context, client_auth);
inner = std::make_unique<StorageS3Source>(
s3_source_builder.need_path,
s3_source_builder.need_file,
s3_source_builder.format,
s3_source_builder.name,
s3_source_builder.sample_block,
s3_source_builder.context,
s3_source_builder.columns,
s3_source_builder.max_block_size,
s3b.need_path, s3b.need_file, s3b.format, s3b.name,
s3b.sample_block, s3b.context, s3b.columns, s3b.max_block_size,
chooseCompressionMethod(client_auth.uri.key, ""),
client_auth.client,
client_auth.uri.bucket,
@ -167,14 +161,10 @@ private:
return true;
}
/// This is used to ask about next task
String initial_query_id;
StorageS3SourceBuilder s3_source_builder;
ClientAuthentificationBuilder cli_builder;
StorageS3SourceBuilder s3b;
ClientAuthentificationBuilder cab;
std::unique_ptr<StorageS3Source> inner;
NextTaskCallback read_task_callback;
};
@ -229,28 +219,18 @@ Pipe StorageS3Distributed::read(
need_file_column = true;
}
StorageS3SourceBuilder s3builder
{
need_path_column,
need_file_column,
format_name,
getName(),
metadata_snapshot->getSampleBlock(),
context,
metadata_snapshot->getColumns(),
max_block_size,
StorageS3SourceBuilder s3builder{
need_path_column, need_file_column,
format_name, getName(),
metadata_snapshot->getSampleBlock(), context,
metadata_snapshot->getColumns(), max_block_size,
compression_method
};
return Pipe(std::make_shared<StorageS3SequentialSource>(
context.getInitialQueryId(),
context.getNextTaskCallback(),
cli_builder,
s3builder
));
context.getInitialQueryId(), context.getNextTaskCallback(), cli_builder, s3builder));
}
/// The code from here and below executes on initiator
/// Calculate the header. This is significant, because some columns could be thrown away in some cases like query with count(*)
@ -269,21 +249,33 @@ Pipe StorageS3Distributed::read(
connections.emplace_back(std::make_shared<Connection>(
node.host_name, node.port, context.getGlobalContext().getCurrentDatabase(),
node.user, node.password, node.cluster, node.cluster_secret,
"StorageS3DistributedInititiator",
Protocol::Compression::Disable,
Protocol::Secure::Disable
"S3DistributedInititiator",
node.compression,
node.secure
));
auto remote_query_executor = std::make_shared<RemoteQueryExecutor>(
*connections.back(), queryToString(query_info.query), header, context, /*throttler=*/nullptr, scalars, Tables(), processed_stage);
pipes.emplace_back(createRemoteSourcePipe(remote_query_executor, false, false, false, false));
pipes.emplace_back(std::make_shared<RemoteSource>(remote_query_executor, false, false));
}
}
metadata_snapshot->check(column_names, getVirtuals(), getStorageID());
return Pipe::unitePipes(std::move(pipes));
}
QueryProcessingStage::Enum StorageS3Distributed::getQueryProcessingStage(const Context & context, QueryProcessingStage::Enum /*to_stage*/, SelectQueryInfo &) const
{
/// Initiator executes query on remote node.
if (context.getClientInfo().query_kind == ClientInfo::QueryKind::INITIAL_QUERY) {
return QueryProcessingStage::Enum::WithMergeableState;
}
/// Follower just reads the data.
return QueryProcessingStage::Enum::FetchColumns;
}
}
#endif

5
src/Storages/StorageS3Distributed.h
View File

@ -46,10 +46,7 @@ public:
size_t /*max_block_size*/,
unsigned /*num_streams*/) override;
QueryProcessingStage::Enum getQueryProcessingStage(const Context &, QueryProcessingStage::Enum /*to_stage*/, SelectQueryInfo &) const override
{
return QueryProcessingStage::Enum::WithMergeableState;
}
QueryProcessingStage::Enum getQueryProcessingStage(const Context & context, QueryProcessingStage::Enum /*to_stage*/, SelectQueryInfo &) const override;
NamesAndTypesList getVirtuals() const override;