Ignite实战之四 集群的管理

在之前的例子中， 我们的程序实现了分布式调用，这是一个最最基础的需求--能跑。随后我们的“野心”会更大，希望能精细化的控制我们的应用跑在哪里，某一个实例上或者某几个实例上。这就需要下层的分布式系统能对集群做精细化的管理，而在Ignite中，对集群的管理大致可以发分为两类

粗放化

• 远程节点分组

IgniteCluster cluster = ignite.cluster();
// Cluster group with remote nodes, i.e. other than this node.
ClusterGroup remoteGroup = cluster.forRemotes();

• client节点分组

IgniteCluster cluster = ignite.cluster();
// All client nodes.
ClusterGroup clientGroup = cluster.forClients();

• server节点分组

IgniteCluster cluster = ignite.cluster();
// All server nodes.
ClusterGroup serverGroup = cluster.forServers();

精细化

• 根据是否部署了特定cache分组

IgniteCluster cluster = ignite.cluster();
// All nodes on which cache with name "myCache" is deployed,
// either in client or server mode.
ClusterGroup cacheGroup = cluster.forCache("myCache");

• 根据部署了特定cache的server节点分组

IgniteCluster cluster = ignite.cluster();
// All data nodes responsible for caching data for "myCache".
ClusterGroup dataGroup = cluster.forDataNodes("myCache");

• 根据可访问特定cache的client节点分组

IgniteCluster cluster = ignite.cluster();
// All client nodes that access "myCache".
ClusterGroup clientGroup = cluster.forClientNodes("myCache");

• 根据节点属性分组
  节点启动时可以配置特定的属性值

<bean class="org.apache.ignite.IgniteConfiguration">
    ...
    <property name="userAttributes">
        <map>
            <entry key="ROLE" value="worker"/>
        </map>
    </property>
    ...
</bean>

之后可以根据属性值进行分组

IgniteCluster cluster = ignite.cluster();
// All nodes with attribute "ROLE" equal to "worker".
ClusterGroup attrGroup = cluster.forAttribute("ROLE", "worker");

如此，我们就可以将应用精确的部署到所需要的分组中了

IgniteCompute compute = ignite.compute(xxxGroup);