Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
2dot5
ClickHouse
提交
43a18b16
C
ClickHouse
项目概览
2dot5
/
ClickHouse
通知
3
Star
0
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
DevOps
流水线
流水线任务
计划
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
C
ClickHouse
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
DevOps
DevOps
流水线
流水线任务
计划
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
流水线任务
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
提交
43a18b16
编写于
6月 15, 2017
作者:
A
alexey-milovidov
提交者:
GitHub
6月 15, 2017
浏览文件
操作
浏览文件
下载
差异文件
Merge pull request #883 from yandex/fix-merge-global-in
Fix GLOBAL IN for Merge table over Distributed tables
上级
f2c1858f
8ee85e64
变更
11
隐藏空白更改
内联
并排
Showing
11 changed file
with
88 addition
and
28 deletion
+88
-28
dbms/src/Interpreters/InJoinSubqueriesPreprocessor.cpp
dbms/src/Interpreters/InJoinSubqueriesPreprocessor.cpp
+0
-3
dbms/src/Interpreters/InterpreterSelectQuery.cpp
dbms/src/Interpreters/InterpreterSelectQuery.cpp
+3
-20
dbms/src/Parsers/ASTSelectQuery.cpp
dbms/src/Parsers/ASTSelectQuery.cpp
+1
-1
dbms/src/Parsers/ASTSelectQuery.h
dbms/src/Parsers/ASTSelectQuery.h
+1
-1
dbms/src/Storages/StorageDistributed.cpp
dbms/src/Storages/StorageDistributed.cpp
+2
-2
dbms/src/Storages/StorageMerge.cpp
dbms/src/Storages/StorageMerge.cpp
+13
-0
dbms/src/Storages/StorageMerge.h
dbms/src/Storages/StorageMerge.h
+2
-0
dbms/tests/integration/helpers/cluster.py
dbms/tests/integration/helpers/cluster.py
+0
-1
dbms/tests/integration/test_merge_table_over_distributed/__init__.py
...integration/test_merge_table_over_distributed/__init__.py
+0
-0
dbms/tests/integration/test_merge_table_over_distributed/configs/remote_servers.xml
...t_merge_table_over_distributed/configs/remote_servers.xml
+18
-0
dbms/tests/integration/test_merge_table_over_distributed/test.py
...sts/integration/test_merge_table_over_distributed/test.py
+48
-0
未找到文件。
dbms/src/Interpreters/InJoinSubqueriesPreprocessor.cpp
浏览文件 @
43a18b16
...
...
@@ -209,9 +209,6 @@ void InJoinSubqueriesPreprocessor::process(ASTSelectQuery * query) const
bool
InJoinSubqueriesPreprocessor
::
hasAtLeastTwoShards
(
const
IStorage
&
table
)
const
{
if
(
!
table
.
isRemote
())
return
false
;
const
StorageDistributed
*
distributed
=
typeid_cast
<
const
StorageDistributed
*>
(
&
table
);
if
(
!
distributed
)
return
false
;
...
...
dbms/src/Interpreters/InterpreterSelectQuery.cpp
浏览文件 @
43a18b16
...
...
@@ -838,24 +838,13 @@ QueryProcessingStage::Enum InterpreterSelectQuery::executeFetchColumns()
if
(
max_streams
>
1
&&
!
is_remote
)
max_streams
*=
settings
.
max_streams_to_max_threads_ratio
;
ASTPtr
actual_query_ptr
;
if
(
storage
->
isRemote
())
{
/// In case of a remote query, we send only SELECT, which will be executed.
actual_query_ptr
=
query
.
cloneFirstSelect
();
}
else
actual_query_ptr
=
query_ptr
;
/// PREWHERE optimization
{
auto
optimize_prewhere
=
[
&
](
auto
&
merge_tree
)
{
const
ASTSelectQuery
&
actual_select
=
typeid_cast
<
const
ASTSelectQuery
&>
(
*
actual_query_ptr
);
/// Try transferring some condition from WHERE to PREWHERE if enabled and viable
if
(
settings
.
optimize_move_to_prewhere
&&
actual_select
.
where_expression
&&
!
actual_select
.
prewhere_expression
&&
!
actual_select
.
final
())
MergeTreeWhereOptimizer
{
actual_
query_ptr
,
context
,
merge_tree
.
getData
(),
required_columns
,
log
};
if
(
settings
.
optimize_move_to_prewhere
&&
query
.
where_expression
&&
!
query
.
prewhere_expression
&&
!
query
.
final
())
MergeTreeWhereOptimizer
{
query_ptr
,
context
,
merge_tree
.
getData
(),
required_columns
,
log
};
};
if
(
const
StorageMergeTree
*
merge_tree
=
typeid_cast
<
const
StorageMergeTree
*>
(
storage
.
get
()))
...
...
@@ -864,8 +853,7 @@ QueryProcessingStage::Enum InterpreterSelectQuery::executeFetchColumns()
optimize_prewhere
(
*
merge_tree
);
}
streams
=
storage
->
read
(
required_columns
,
actual_query_ptr
,
context
,
from_stage
,
max_block_size
,
max_streams
);
streams
=
storage
->
read
(
required_columns
,
query_ptr
,
context
,
from_stage
,
max_block_size
,
max_streams
);
if
(
alias_actions
)
{
...
...
@@ -1316,11 +1304,6 @@ void InterpreterSelectQuery::executeLimit()
void
InterpreterSelectQuery
::
executeSubqueriesInSetsAndJoins
(
SubqueriesForSets
&
subqueries_for_sets
)
{
/// If the query is not distributed, then remove the creation of temporary tables from subqueries (intended for sending to remote servers).
if
(
!
(
storage
&&
storage
->
isRemote
()))
for
(
auto
&
elem
:
subqueries_for_sets
)
elem
.
second
.
table
.
reset
();
const
Settings
&
settings
=
context
.
getSettingsRef
();
executeUnion
();
...
...
dbms/src/Parsers/ASTSelectQuery.cpp
浏览文件 @
43a18b16
...
...
@@ -173,7 +173,7 @@ ASTPtr ASTSelectQuery::clone() const
return
ptr
;
}
ASTPtr
ASTSelectQuery
::
cloneFirstSelect
()
const
std
::
shared_ptr
<
ASTSelectQuery
>
ASTSelectQuery
::
cloneFirstSelect
()
const
{
auto
res
=
cloneImpl
(
false
);
res
->
prev_union_all
=
nullptr
;
...
...
dbms/src/Parsers/ASTSelectQuery.h
浏览文件 @
43a18b16
...
...
@@ -39,7 +39,7 @@ public:
ASTPtr
clone
()
const
override
;
/// Get a deep copy of the first SELECT query tree.
ASTPtr
cloneFirstSelect
()
const
;
std
::
shared_ptr
<
ASTSelectQuery
>
cloneFirstSelect
()
const
;
private:
std
::
shared_ptr
<
ASTSelectQuery
>
cloneImpl
(
bool
traverse_union_all
)
const
;
...
...
dbms/src/Storages/StorageDistributed.cpp
浏览文件 @
43a18b16
...
...
@@ -60,8 +60,8 @@ namespace
/// Creates a copy of query, changes database and table names.
ASTPtr
rewriteSelectQuery
(
const
ASTPtr
&
query
,
const
std
::
string
&
database
,
const
std
::
string
&
table
)
{
auto
modified_query_ast
=
query
->
clone
();
typeid_cast
<
ASTSelectQuery
&>
(
*
modified_query_ast
).
replaceDatabaseAndTable
(
database
,
table
);
auto
modified_query_ast
=
typeid_cast
<
const
ASTSelectQuery
&>
(
*
query
).
cloneFirstSelect
();
modified_query_ast
->
replaceDatabaseAndTable
(
database
,
table
);
return
modified_query_ast
;
}
...
...
dbms/src/Storages/StorageMerge.cpp
浏览文件 @
43a18b16
...
...
@@ -50,6 +50,19 @@ StorageMerge::StorageMerge(
{
}
bool
StorageMerge
::
isRemote
()
const
{
auto
database
=
context
.
getDatabase
(
source_database
);
auto
iterator
=
database
->
getIterator
();
while
(
iterator
->
isValid
())
{
if
(
table_name_regexp
.
match
(
iterator
->
name
())
&&
iterator
->
table
()
->
isRemote
())
return
true
;
}
return
false
;
}
NameAndTypePair
StorageMerge
::
getColumn
(
const
String
&
column_name
)
const
{
auto
type
=
VirtualColumnFactory
::
tryGetType
(
column_name
);
...
...
dbms/src/Storages/StorageMerge.h
浏览文件 @
43a18b16
...
...
@@ -20,6 +20,8 @@ public:
std
::
string
getName
()
const
override
{
return
"Merge"
;
}
std
::
string
getTableName
()
const
override
{
return
name
;
}
bool
isRemote
()
const
override
;
/// The check is delayed to the read method. It checks the support of the tables used.
bool
supportsSampling
()
const
override
{
return
true
;
}
bool
supportsPrewhere
()
const
override
{
return
true
;
}
...
...
dbms/tests/integration/helpers/cluster.py
浏览文件 @
43a18b16
...
...
@@ -113,7 +113,6 @@ class ClickHouseCluster:
for
instance
in
self
.
instances
.
values
():
instance
.
docker_client
=
None
instance
.
docker_id
=
None
instance
.
ip_address
=
None
instance
.
client
=
None
...
...
dbms/tests/integration/test_merge_table_over_distributed/__init__.py
0 → 100644
浏览文件 @
43a18b16
dbms/tests/integration/test_merge_table_over_distributed/configs/remote_servers.xml
0 → 100644
浏览文件 @
43a18b16
<yandex>
<remote_servers>
<test_cluster>
<shard>
<replica>
<host>
node1
</host>
<port>
9000
</port>
</replica>
</shard>
<shard>
<replica>
<host>
node2
</host>
<port>
9000
</port>
</replica>
</shard>
</test_cluster>
</remote_servers>
</yandex>
dbms/tests/integration/test_merge_table_over_distributed/test.py
0 → 100644
浏览文件 @
43a18b16
from
contextlib
import
contextmanager
import
pytest
from
helpers.cluster
import
ClickHouseCluster
cluster
=
ClickHouseCluster
(
__file__
)
node1
=
cluster
.
add_instance
(
'node1'
,
main_configs
=
[
'configs/remote_servers.xml'
])
node2
=
cluster
.
add_instance
(
'node2'
,
main_configs
=
[
'configs/remote_servers.xml'
])
@
pytest
.
fixture
(
scope
=
"module"
)
def
started_cluster
():
try
:
cluster
.
start
()
for
node
in
(
node1
,
node2
):
node
.
query
(
'''
CREATE TABLE local_table(id UInt32, val String) ENGINE = TinyLog;
'''
)
node1
.
query
(
"INSERT INTO local_table VALUES (1, 'node1')"
)
node2
.
query
(
"INSERT INTO local_table VALUES (2, 'node2')"
)
node1
.
query
(
'''
CREATE TABLE distributed_table(id UInt32, val String) ENGINE = Distributed(test_cluster, default, local_table);
CREATE TABLE merge_table(id UInt32, val String) ENGINE = Merge(default, '^distributed_table')
'''
)
yield
cluster
finally
:
cluster
.
shutdown
()
def
test_global_in
(
started_cluster
):
assert
node1
.
query
(
"SELECT val FROM distributed_table WHERE id GLOBAL IN (SELECT toUInt32(3 - id) FROM local_table)"
).
rstrip
()
\
==
'node2'
assert
node1
.
query
(
"SELECT val FROM merge_table WHERE id GLOBAL IN (SELECT toUInt32(3 - id) FROM local_table)"
).
rstrip
()
\
==
'node2'
if
__name__
==
'__main__'
:
with
contextmanager
(
started_cluster
)()
as
cluster
:
for
name
,
instance
in
cluster
.
instances
.
items
():
print
name
,
instance
.
ip_address
raw_input
(
"Cluster created, press any key to destroy..."
)
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录