mirror of
https://github.com/didi/KnowStreaming.git
synced 2026-01-07 15:12:14 +08:00
99 lines
1.1 KiB
Markdown
99 lines
1.1 KiB
Markdown
|
||

|
||
|
||
|
||
|
||
|
||
# 健康巡检
|
||
|
||
## 1、前言
|
||
|
||
|
||
|
||
---
|
||
|
||
## 2、已有巡检
|
||
|
||
### 2.1、Cluster健康巡检(1个)
|
||
|
||
#### 2.1.1、集群Controller数错误
|
||
|
||
**说明**
|
||
|
||
- 集群Controller数不等于1,表明集群集群无Controller或者出现了多个Controller,该
|
||
|
||
|
||
**配置**
|
||
|
||
---
|
||
|
||
### 2.2、Broker健康巡检(2个)
|
||
|
||
#### 2.2.1、Broker-RequestQueueSize被打满
|
||
|
||
**说明**
|
||
|
||
- Broker的RequestQueueSize,被打满;
|
||
|
||
|
||
**配置**
|
||
|
||
---
|
||
|
||
|
||
#### 2.2.2、Broker-NetworkProcessorAvgIdle过低
|
||
|
||
**说明**
|
||
|
||
- Broker的NetworkProcessorAvgIdle指标,当前过低;
|
||
|
||
|
||
**配置**
|
||
|
||
---
|
||
|
||
### 2.3、Topic健康巡检(2个)
|
||
|
||
|
||
#### 2.3.1、Topic 无Leader数
|
||
|
||
**说明**
|
||
|
||
- 当前Topic的无Leader分区数超过一定值;
|
||
|
||
|
||
**配置**
|
||
|
||
|
||
#### 2.3.1、Topic 长期处于未同步状态
|
||
|
||
**说明**
|
||
|
||
- 指定的一段时间内,Topic一直处于未同步的状态;
|
||
|
||
|
||
**配置**
|
||
|
||
---
|
||
|
||
### 2.4、Group健康巡检(1个)
|
||
|
||
|
||
#### 2.4.1、Group Re-Balance太频繁
|
||
|
||
**说明**
|
||
|
||
- 指定的一段时间内,Group Re-Balance的次数是否过多;
|
||
|
||
|
||
**配置**
|
||
|
||
|
||
|
||
---
|
||
|
||
## 3、自定义增强
|
||
|
||
如何增加想要的巡检?
|
||
|