Yarn Label 调度

喜欢ヅ旅行 2022-05-15 01:46 239阅读 0赞

### 1. `Yarn Label`标签创建 ###

当重启`ResourceManager`时，可以保证`Yarn Label`标签在服务重启之后依然可用，需要在`HDFS`上指定 Yarn Label的存储路径，具体操作如下：

1.  在`HDFS`上创建目录，用于存放`Yarn Label`标签数据（注意创建的目录`Yarn`要可以访问），如下命令可能会被使用到

hdfs dfs -mkdir -p /yarn/node-labels
    hdfs dfs -chown -R yarn:yarn /yarn/node-labels
    hdfs dfs -chmod -R 700 /yarn/node-labels

1.  启动集群标签功能

<property>
        <name>yarn.node-labels.enabled</name>
        <value>true</value>
    </property>
    <property>
        <name>yarn.node-labels.fs-store.root-dir</name>
        <value>hdfs://<host>:<port>/<absolute_path_to_node_label_directory></value>
    </property>
    <property>
        <name>yarn.resourcemanager.scheduler.class</name>
        <value>org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler</value>
    </property>

> 重启`ResourceManager`配置生效

yarn-daemon.sh stop  resourcemanager
    yarn-daemon.sh start resourcemanager

1.  为`yarn`创建**集群**标签

yarn rmadmin -addToClusterNodeLabels "<label1>(exclusive=<true|false>),<label2>(exclusive=<true|false>)"

> 注：`exclusive`并不是必须的参数，默认值为true  
> 4. 删除**集群**标签

yarn rmadmin -removeFromClusterNodeLabels "<label1>,<label2>"

> 注：如果某个Label已经关联Queue，那么你无法从集群中将其移除

### 2. 为集群节点创建Label标签 ###

1.  为**节点**关联标签

yarn rmadmin -replaceLabelsOnNode "<node1>:<port>=<label1> <node2>:<port>=<label2>"

> 注：这里用于关联的集群标签必须是之前定义好的，这一步是将集群标签与节点关联（为节点打标签）  
> 2. 删除**节点**标签 (将节点的标签置为空)

yarn rmadmin -replaceLabelsOnNode "<node_name>"

### 3. 创建Queue并关联标签 ###

<configuration>
        
        <property>
            <name>yarn.scheduler.capacity.root.queues</name>
            <value>pro,dev,def</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.maximum-capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.accessible-node-labels</name>
            <value>*</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.accessible-node-labels.cpu.capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.accessible-node-labels.cpu.maximum-capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.accessible-node-labels.mem.capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.accessible-node-labels.mem.maximum-capacity</name>
            <value>100</value>
        </property>
    
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.capacity</name>
            <value>40</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.maximum-capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.accessible-node-labels</name>
            <value>cpu,mem</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.default-node-label-expression</name>
            <value>cpu</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.accessible-node-labels.cpu.capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.accessible-node-labels.cpu.maximum-capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.accessible-node-labels.mem.capacity</name>
            <value>50</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.pro.accessible-node-labels.mem.maximum-capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.dev.capacity</name>
            <value>30</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.dev.maximum-capacity</name>
            <value>100</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.dev.accessible-node-labels</name>
            <value>mem</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.dev.accessible-node-labels.mem.capacity</name>
            <value>50</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.dev.accessible-node-labels.mem.maximum-capacity</name>
            <value>100</value>
        </property>
    
        
        <property>
            <name>yarn.scheduler.capacity.root.def.capacity</name>
            <value>30</value>
        </property>
        
        <property>
            <name>yarn.scheduler.capacity.root.def.maximum-capacity</name>
            <value>100</value>
        </property>
    </configuration>

> 注：**在配置完capacity-scheduler.xml之后，需要刷新下队列**，很重要，在删除集群标签时也要修改该配置文件，然后刷新列表

yarn rmadmin -refreshQueues