This is an automated email from the ASF dual-hosted git repository.
kassiez pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/doris-website.git
The following commit(s) were added to refs/heads/master by this push:
new 5ecfcfcd52e [fix]Fixed formatting issues in Analysis Tools
documentation (#2185)
5ecfcfcd52e is described below
commit 5ecfcfcd52e7d8a49af4e4b8ffda0d55303c539c
Author: wudongliang <[email protected]>
AuthorDate: Mon Mar 31 20:10:48 2025 +0800
[fix]Fixed formatting issues in Analysis Tools documentation (#2185)
## Versions
- [x] dev
- [x] 3.0
- [x] 2.1
- [ ] 2.0
## Languages
- [x] Chinese
- [x] English
## Docs Checklist
- [ ] Checked by AI
- [ ] Test Cases Built
---
.../performance-tuning-overview/analysis-tools.md | 29 +++++++++++--
.../performance-tuning-overview/analysis-tools.md | 48 +++++++++++-----------
.../performance-tuning-overview/analysis-tools.md | 46 ++++++++++-----------
.../performance-tuning-overview/analysis-tools.md | 48 +++++++++++-----------
.../performance-tuning-overview/analysis-tools.md | 29 +++++++++++--
.../performance-tuning-overview/analysis-tools.md | 29 +++++++++++--
6 files changed, 146 insertions(+), 83 deletions(-)
diff --git
a/docs/query-acceleration/performance-tuning-overview/analysis-tools.md
b/docs/query-acceleration/performance-tuning-overview/analysis-tools.md
index 4c96db5bd91..12eec97564f 100644
--- a/docs/query-acceleration/performance-tuning-overview/analysis-tools.md
+++ b/docs/query-acceleration/performance-tuning-overview/analysis-tools.md
@@ -128,10 +128,31 @@ In this profile, for instance, LocalBytesReceived is a
metric specific to the ex
### PipelineTask Execution Time
In Doris, a PipelineTask consists of multiple operators. When analyzing the
execution time of a PipelineTask, several key aspects need to be focused on:
-
-1. ExecuteTime: 1.656ms (The actual execution time of the entire PipelineTask,
which is approximately the sum of the ExecTime of all operators within the
task).
-2. WaitWorkerTime: 63.868us (The time the task waits for an execution worker.
When the task is in a runnable state, it waits for an available worker to
execute it, and this duration primarily depends on the cluster load).
-3. Time Waiting for Execution Dependencies: 10.495ms (WaitForBroadcastBuffer +
WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time). The time a task waits
for execution dependencies is the sum of the waiting times for these
dependencies.
+1. ExecuteTime: The actual execution time of the entire PipelineTask, which is
approximately equal to the sum of the ExecTime of all operators in this task
+2. WaitWorkerTime: The time that a task waits for a worker to execute. When a
task is in the runnable state, it has to wait for an idle worker to execute it.
The time it takes depends mainly on the cluster load.
+3. Waiting time for executing dependencies: A task can be executed only when
all the dependencies of each operator meet the execution conditions, and the
time a task waits for executing dependencies is the sum of the waiting times of
these dependencies. For example, simplifying one of the tasks in this example:
+
+ ```sql
+ PipelineTask (index=1):(ExecTime: 4.773ms)
+ - ExecuteTime: 1.656ms
+ - CloseTime: 90.402us
+ - GetBlockTime: 11.235us
+ - OpenTime: 1.448ms
+ - PrepareTime: 1.555ms
+ - SinkTime: 14.228us
+ - WaitWorkerTime: 63.868us
+ DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
+ - WaitForDependencyTime: 0ns
+ - WaitForBroadcastBuffer: 0ns
+ - WaitForRpcBufferQueue: 0ns
+ AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
+ - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
+ ```
+ This task includes two operators (DATA_STREAM_SINK_OPERATOR -
AGGREGATION_OPERATOR), of which DATA_STREAM_SINK_OPERATOR has two dependencies
(WaitForBroadcastBuffer and WaitForRpcBufferQueue), and AGGREGATION_OPERATOR
has one dependency (AGGREGATION_OPERATOR_DEPENDENCY), so the time consumption
of the current task is distributed as follows:
+
+ 1. ExecuteTime: 1.656ms (The actual execution time of the entire
PipelineTask, which is approximately the sum of the ExecTime of all operators
within the task).
+ 2. WaitWorkerTime: 63.868us (The time the task waits for an execution
worker. When the task is in a runnable state, it waits for an available worker
to execute it, and this duration primarily depends on the cluster load).
+ 3. Time Waiting for Execution Dependencies: 10.495ms
(WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time). The time a task waits
for execution dependencies is the sum of the waiting times for these
dependencies.
For cases of using Profile for execution-level tuning, please refer to the
[Tuning
Execution](../tuning/tuning-execution/adjustment-of-runtimefilter-wait-time.md)
section.
diff --git
a/i18n/zh-CN/docusaurus-plugin-content-docs/current/query-acceleration/performance-tuning-overview/analysis-tools.md
b/i18n/zh-CN/docusaurus-plugin-content-docs/current/query-acceleration/performance-tuning-overview/analysis-tools.md
index cf6b34beaa5..c963b8d83a7 100644
---
a/i18n/zh-CN/docusaurus-plugin-content-docs/current/query-acceleration/performance-tuning-overview/analysis-tools.md
+++
b/i18n/zh-CN/docusaurus-plugin-content-docs/current/query-acceleration/performance-tuning-overview/analysis-tools.md
@@ -1,7 +1,7 @@
---
{
- "title": "分析工具",
- "language": "zh-CN"
+ "title": "分析工具",
+ "language": "zh-CN"
}
---
@@ -133,28 +133,28 @@ EXCHANGE_OPERATOR (id=4):(ExecTime: 706.351us)
2. WaitWorkerTime：task 等待执行 worker 的时间。当 task 处于 runnable 状态时，他要等待一个空闲 worker
来执行，这个耗时主要取决于集群负载。
3. 等待执行依赖的时间：一个 task 可以执行的依赖条件是每个 operator 的 dependency 全部满足执行条件，而 task
等待执行依赖的时间就是将这些依赖的等待时间相加。例如简化这个例子中的其中一个 task：
-```sql
-PipelineTask (index=1):(ExecTime: 4.773ms)
- - ExecuteTime: 1.656ms
- - CloseTime: 90.402us
- - GetBlockTime: 11.235us
- - OpenTime: 1.448ms
- - PrepareTime: 1.555ms
- - SinkTime: 14.228us
- - WaitWorkerTime: 63.868us
- DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
- - WaitForDependencyTime: 0ns
- - WaitForBroadcastBuffer: 0ns
- - WaitForRpcBufferQueue: 0ns
- AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
- - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
-```
-
-这个 task 包含了（DATA_STREAM_SINK_OPERATOR - AGGREGATION_OPERATOR）两个 operator，其中
DATA_STREAM_SINK_OPERATOR 有两个依赖（WaitForBroadcastBuffer 和
WaitForRpcBufferQueue），AGGREGATION_OPERATOR
有一个依赖（AGGREGATION_OPERATOR_DEPENDENCY），所以当前 task 的耗时分布如下：
-
-1. 执行总时间：1.656ms（约等于两个 operator 的 ExecTime 总和）
-2. 等待 Worker 的时间：63.868us（说明当前集群负载不高，task 就绪以后立即就有 worker 来执行）
-3. 等待执行依赖的时间（WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time）：10.495ms。当前 task 的所有
dependency 相加得到总的等待时间。
+ ```sql
+ PipelineTask (index=1):(ExecTime: 4.773ms)
+ - ExecuteTime: 1.656ms
+ - CloseTime: 90.402us
+ - GetBlockTime: 11.235us
+ - OpenTime: 1.448ms
+ - PrepareTime: 1.555ms
+ - SinkTime: 14.228us
+ - WaitWorkerTime: 63.868us
+ DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
+ - WaitForDependencyTime: 0ns
+ - WaitForBroadcastBuffer: 0ns
+ - WaitForRpcBufferQueue: 0ns
+ AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
+ - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
+ ```
+
+ 这个 task 包含了（DATA_STREAM_SINK_OPERATOR - AGGREGATION_OPERATOR）两个 operator，其中
DATA_STREAM_SINK_OPERATOR 有两个依赖（WaitForBroadcastBuffer 和
WaitForRpcBufferQueue），AGGREGATION_OPERATOR
有一个依赖（AGGREGATION_OPERATOR_DEPENDENCY），所以当前 task 的耗时分布如下：
+
+ 1. 执行总时间：1.656ms（约等于两个 operator 的 ExecTime 总和）
+ 2. 等待 Worker 的时间：63.868us（说明当前集群负载不高，task 就绪以后立即就有 worker 来执行）
+ 3. 等待执行依赖的时间（WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time）：10.495ms。当前 task 的所有
dependency 相加得到总的等待时间。
使用 Profile
进行执行层调优的案例详见[执行调优](../tuning/tuning-execution/adjustment-of-runtimefilter-wait-time.md)章节。
diff --git
a/i18n/zh-CN/docusaurus-plugin-content-docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
b/i18n/zh-CN/docusaurus-plugin-content-docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
index cf6b34beaa5..3794b558521 100644
---
a/i18n/zh-CN/docusaurus-plugin-content-docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
+++
b/i18n/zh-CN/docusaurus-plugin-content-docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
@@ -132,29 +132,29 @@ EXCHANGE_OPERATOR (id=4):(ExecTime: 706.351us)
1. ExecuteTime：整个 PipelineTask 的实际执行时间，约等于这个 task 中所有 operator 的 ExecTime 相加
2. WaitWorkerTime：task 等待执行 worker 的时间。当 task 处于 runnable 状态时，他要等待一个空闲 worker
来执行，这个耗时主要取决于集群负载。
3. 等待执行依赖的时间：一个 task 可以执行的依赖条件是每个 operator 的 dependency 全部满足执行条件，而 task
等待执行依赖的时间就是将这些依赖的等待时间相加。例如简化这个例子中的其中一个 task：
-
-```sql
-PipelineTask (index=1):(ExecTime: 4.773ms)
- - ExecuteTime: 1.656ms
- - CloseTime: 90.402us
- - GetBlockTime: 11.235us
- - OpenTime: 1.448ms
- - PrepareTime: 1.555ms
- - SinkTime: 14.228us
- - WaitWorkerTime: 63.868us
- DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
- - WaitForDependencyTime: 0ns
- - WaitForBroadcastBuffer: 0ns
- - WaitForRpcBufferQueue: 0ns
- AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
- - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
-```
-
-这个 task 包含了（DATA_STREAM_SINK_OPERATOR - AGGREGATION_OPERATOR）两个 operator，其中
DATA_STREAM_SINK_OPERATOR 有两个依赖（WaitForBroadcastBuffer 和
WaitForRpcBufferQueue），AGGREGATION_OPERATOR
有一个依赖（AGGREGATION_OPERATOR_DEPENDENCY），所以当前 task 的耗时分布如下：
-
-1. 执行总时间：1.656ms（约等于两个 operator 的 ExecTime 总和）
-2. 等待 Worker 的时间：63.868us（说明当前集群负载不高，task 就绪以后立即就有 worker 来执行）
-3. 等待执行依赖的时间（WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time）：10.495ms。当前 task 的所有
dependency 相加得到总的等待时间。
+
+ ```sql
+ PipelineTask (index=1):(ExecTime: 4.773ms)
+ - ExecuteTime: 1.656ms
+ - CloseTime: 90.402us
+ - GetBlockTime: 11.235us
+ - OpenTime: 1.448ms
+ - PrepareTime: 1.555ms
+ - SinkTime: 14.228us
+ - WaitWorkerTime: 63.868us
+ DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
+ - WaitForDependencyTime: 0ns
+ - WaitForBroadcastBuffer: 0ns
+ - WaitForRpcBufferQueue: 0ns
+ AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
+ - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
+ ```
+
+ 这个 task 包含了（DATA_STREAM_SINK_OPERATOR - AGGREGATION_OPERATOR）两个
operator，其中 DATA_STREAM_SINK_OPERATOR 有两个依赖（WaitForBroadcastBuffer 和
WaitForRpcBufferQueue），AGGREGATION_OPERATOR
有一个依赖（AGGREGATION_OPERATOR_DEPENDENCY），所以当前 task 的耗时分布如下：
+
+ 1. 执行总时间：1.656ms（约等于两个 operator 的 ExecTime 总和）
+ 2. 等待 Worker 的时间：63.868us（说明当前集群负载不高，task 就绪以后立即就有 worker 来执行）
+ 3. 等待执行依赖的时间（WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time）：10.495ms。当前 task 的所有
dependency 相加得到总的等待时间。
使用 Profile
进行执行层调优的案例详见[执行调优](../tuning/tuning-execution/adjustment-of-runtimefilter-wait-time.md)章节。
diff --git
a/i18n/zh-CN/docusaurus-plugin-content-docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
b/i18n/zh-CN/docusaurus-plugin-content-docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
index cf6b34beaa5..c963b8d83a7 100644
---
a/i18n/zh-CN/docusaurus-plugin-content-docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
+++
b/i18n/zh-CN/docusaurus-plugin-content-docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
@@ -1,7 +1,7 @@
---
{
- "title": "分析工具",
- "language": "zh-CN"
+ "title": "分析工具",
+ "language": "zh-CN"
}
---
@@ -133,28 +133,28 @@ EXCHANGE_OPERATOR (id=4):(ExecTime: 706.351us)
2. WaitWorkerTime：task 等待执行 worker 的时间。当 task 处于 runnable 状态时，他要等待一个空闲 worker
来执行，这个耗时主要取决于集群负载。
3. 等待执行依赖的时间：一个 task 可以执行的依赖条件是每个 operator 的 dependency 全部满足执行条件，而 task
等待执行依赖的时间就是将这些依赖的等待时间相加。例如简化这个例子中的其中一个 task：
-```sql
-PipelineTask (index=1):(ExecTime: 4.773ms)
- - ExecuteTime: 1.656ms
- - CloseTime: 90.402us
- - GetBlockTime: 11.235us
- - OpenTime: 1.448ms
- - PrepareTime: 1.555ms
- - SinkTime: 14.228us
- - WaitWorkerTime: 63.868us
- DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
- - WaitForDependencyTime: 0ns
- - WaitForBroadcastBuffer: 0ns
- - WaitForRpcBufferQueue: 0ns
- AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
- - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
-```
-
-这个 task 包含了（DATA_STREAM_SINK_OPERATOR - AGGREGATION_OPERATOR）两个 operator，其中
DATA_STREAM_SINK_OPERATOR 有两个依赖（WaitForBroadcastBuffer 和
WaitForRpcBufferQueue），AGGREGATION_OPERATOR
有一个依赖（AGGREGATION_OPERATOR_DEPENDENCY），所以当前 task 的耗时分布如下：
-
-1. 执行总时间：1.656ms（约等于两个 operator 的 ExecTime 总和）
-2. 等待 Worker 的时间：63.868us（说明当前集群负载不高，task 就绪以后立即就有 worker 来执行）
-3. 等待执行依赖的时间（WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time）：10.495ms。当前 task 的所有
dependency 相加得到总的等待时间。
+ ```sql
+ PipelineTask (index=1):(ExecTime: 4.773ms)
+ - ExecuteTime: 1.656ms
+ - CloseTime: 90.402us
+ - GetBlockTime: 11.235us
+ - OpenTime: 1.448ms
+ - PrepareTime: 1.555ms
+ - SinkTime: 14.228us
+ - WaitWorkerTime: 63.868us
+ DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
+ - WaitForDependencyTime: 0ns
+ - WaitForBroadcastBuffer: 0ns
+ - WaitForRpcBufferQueue: 0ns
+ AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
+ - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
+ ```
+
+ 这个 task 包含了（DATA_STREAM_SINK_OPERATOR - AGGREGATION_OPERATOR）两个 operator，其中
DATA_STREAM_SINK_OPERATOR 有两个依赖（WaitForBroadcastBuffer 和
WaitForRpcBufferQueue），AGGREGATION_OPERATOR
有一个依赖（AGGREGATION_OPERATOR_DEPENDENCY），所以当前 task 的耗时分布如下：
+
+ 1. 执行总时间：1.656ms（约等于两个 operator 的 ExecTime 总和）
+ 2. 等待 Worker 的时间：63.868us（说明当前集群负载不高，task 就绪以后立即就有 worker 来执行）
+ 3. 等待执行依赖的时间（WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time）：10.495ms。当前 task 的所有
dependency 相加得到总的等待时间。
使用 Profile
进行执行层调优的案例详见[执行调优](../tuning/tuning-execution/adjustment-of-runtimefilter-wait-time.md)章节。
diff --git
a/versioned_docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
b/versioned_docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
index 4c96db5bd91..6b660270209 100644
---
a/versioned_docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
+++
b/versioned_docs/version-2.1/query-acceleration/performance-tuning-overview/analysis-tools.md
@@ -128,10 +128,31 @@ In this profile, for instance, LocalBytesReceived is a
metric specific to the ex
### PipelineTask Execution Time
In Doris, a PipelineTask consists of multiple operators. When analyzing the
execution time of a PipelineTask, several key aspects need to be focused on:
-
-1. ExecuteTime: 1.656ms (The actual execution time of the entire PipelineTask,
which is approximately the sum of the ExecTime of all operators within the
task).
-2. WaitWorkerTime: 63.868us (The time the task waits for an execution worker.
When the task is in a runnable state, it waits for an available worker to
execute it, and this duration primarily depends on the cluster load).
-3. Time Waiting for Execution Dependencies: 10.495ms (WaitForBroadcastBuffer +
WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time). The time a task waits
for execution dependencies is the sum of the waiting times for these
dependencies.
+1. ExecuteTime: The actual execution time of the entire PipelineTask, which is
approximately equal to the sum of the ExecTime of all operators in this task
+2. WaitWorkerTime: The time that a task waits for a worker to execute. When a
task is in the runnable state, it has to wait for an idle worker to execute it.
The time it takes depends mainly on the cluster load.
+3. Waiting time for executing dependencies: A task can be executed only when
all the dependencies of each operator meet the execution conditions, and the
time a task waits for executing dependencies is the sum of the waiting times of
these dependencies. For example, simplifying one of the tasks in this example:
+
+ ```sql
+ PipelineTask (index=1):(ExecTime: 4.773ms)
+ - ExecuteTime: 1.656ms
+ - CloseTime: 90.402us
+ - GetBlockTime: 11.235us
+ - OpenTime: 1.448ms
+ - PrepareTime: 1.555ms
+ - SinkTime: 14.228us
+ - WaitWorkerTime: 63.868us
+ DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
+ - WaitForDependencyTime: 0ns
+ - WaitForBroadcastBuffer: 0ns
+ - WaitForRpcBufferQueue: 0ns
+ AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
+ - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
+ ```
+ This task includes two operators (DATA_STREAM_SINK_OPERATOR -
AGGREGATION_OPERATOR), of which DATA_STREAM_SINK_OPERATOR has two dependencies
(WaitForBroadcastBuffer and WaitForRpcBufferQueue), and AGGREGATION_OPERATOR
has one dependency (AGGREGATION_OPERATOR_DEPENDENCY), so the time consumption
of the current task is distributed as follows:
+
+ 1. ExecuteTime: 1.656ms (The actual execution time of the entire
PipelineTask, which is approximately the sum of the ExecTime of all operators
within the task).
+ 2. WaitWorkerTime: 63.868us (The time the task waits for an execution
worker. When the task is in a runnable state, it waits for an available worker
to execute it, and this duration primarily depends on the cluster load).
+ 3. Time Waiting for Execution Dependencies: 10.495ms
(WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time). The time a task waits
for execution dependencies is the sum of the waiting times for these
dependencies.
For cases of using Profile for execution-level tuning, please refer to the
[Tuning
Execution](../tuning/tuning-execution/adjustment-of-runtimefilter-wait-time.md)
section.
diff --git
a/versioned_docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
b/versioned_docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
index 4c96db5bd91..6b660270209 100644
---
a/versioned_docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
+++
b/versioned_docs/version-3.0/query-acceleration/performance-tuning-overview/analysis-tools.md
@@ -128,10 +128,31 @@ In this profile, for instance, LocalBytesReceived is a
metric specific to the ex
### PipelineTask Execution Time
In Doris, a PipelineTask consists of multiple operators. When analyzing the
execution time of a PipelineTask, several key aspects need to be focused on:
-
-1. ExecuteTime: 1.656ms (The actual execution time of the entire PipelineTask,
which is approximately the sum of the ExecTime of all operators within the
task).
-2. WaitWorkerTime: 63.868us (The time the task waits for an execution worker.
When the task is in a runnable state, it waits for an available worker to
execute it, and this duration primarily depends on the cluster load).
-3. Time Waiting for Execution Dependencies: 10.495ms (WaitForBroadcastBuffer +
WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time). The time a task waits
for execution dependencies is the sum of the waiting times for these
dependencies.
+1. ExecuteTime: The actual execution time of the entire PipelineTask, which is
approximately equal to the sum of the ExecTime of all operators in this task
+2. WaitWorkerTime: The time that a task waits for a worker to execute. When a
task is in the runnable state, it has to wait for an idle worker to execute it.
The time it takes depends mainly on the cluster load.
+3. Waiting time for executing dependencies: A task can be executed only when
all the dependencies of each operator meet the execution conditions, and the
time a task waits for executing dependencies is the sum of the waiting times of
these dependencies. For example, simplifying one of the tasks in this example:
+
+ ```sql
+ PipelineTask (index=1):(ExecTime: 4.773ms)
+ - ExecuteTime: 1.656ms
+ - CloseTime: 90.402us
+ - GetBlockTime: 11.235us
+ - OpenTime: 1.448ms
+ - PrepareTime: 1.555ms
+ - SinkTime: 14.228us
+ - WaitWorkerTime: 63.868us
+ DATA_STREAM_SINK_OPERATOR (id=8,dst_id=8):(ExecTime: 1.688ms)
+ - WaitForDependencyTime: 0ns
+ - WaitForBroadcastBuffer: 0ns
+ - WaitForRpcBufferQueue: 0ns
+ AGGREGATION_OPERATOR (id=7 , nereids_id=648):(ExecTime: 398.12us)
+ - WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time: 10.495ms
+ ```
+ This task includes two operators (DATA_STREAM_SINK_OPERATOR -
AGGREGATION_OPERATOR), of which DATA_STREAM_SINK_OPERATOR has two dependencies
(WaitForBroadcastBuffer and WaitForRpcBufferQueue), and AGGREGATION_OPERATOR
has one dependency (AGGREGATION_OPERATOR_DEPENDENCY), so the time consumption
of the current task is distributed as follows:
+
+ 1. ExecuteTime: 1.656ms (The actual execution time of the entire
PipelineTask, which is approximately the sum of the ExecTime of all operators
within the task).
+ 2. WaitWorkerTime: 63.868us (The time the task waits for an execution
worker. When the task is in a runnable state, it waits for an available worker
to execute it, and this duration primarily depends on the cluster load).
+ 3. Time Waiting for Execution Dependencies: 10.495ms
(WaitForBroadcastBuffer + WaitForRpcBufferQueue +
WaitForDependency[AGGREGATION_OPERATOR_DEPENDENCY]Time). The time a task waits
for execution dependencies is the sum of the waiting times for these
dependencies.
For cases of using Profile for execution-level tuning, please refer to the
[Tuning
Execution](../tuning/tuning-execution/adjustment-of-runtimefilter-wait-time.md)
section.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
