@InterfaceAudience.Private public final class ClusterConnectionFactory extends Object
AsyncClusterConnection.| Modifier and Type | Field and Description |
|---|---|
static String |
HBASE_SERVER_CLUSTER_CONNECTION_IMPL |
| Modifier and Type | Method and Description |
|---|---|
static AsyncClusterConnection |
createAsyncClusterConnection(org.apache.hadoop.conf.Configuration conf,
SocketAddress localAddress,
User user)
Create a new
AsyncClusterConnection instance. |
static AsyncClusterConnection |
createAsyncClusterConnection(ConnectionRegistryEndpoint endpoint,
org.apache.hadoop.conf.Configuration conf,
SocketAddress localAddress,
User user)
Create a new
AsyncClusterConnection instance to be used at server side where we have a
ConnectionRegistryEndpoint. |
public static final String HBASE_SERVER_CLUSTER_CONNECTION_IMPL
public static AsyncClusterConnection createAsyncClusterConnection(org.apache.hadoop.conf.Configuration conf, SocketAddress localAddress, User user) throws IOException
AsyncClusterConnection instance.
Unlike what we have done in ConnectionFactory, here we just return an
AsyncClusterConnection instead of a CompletableFuture,
which means this method could block on fetching the cluster id. This is just used to simplify
the implementation, as when starting new region servers, we do not need to be event-driven. Can
change later if we want a CompletableFuture here.IOExceptionpublic static AsyncClusterConnection createAsyncClusterConnection(ConnectionRegistryEndpoint endpoint, org.apache.hadoop.conf.Configuration conf, SocketAddress localAddress, User user) throws IOException
AsyncClusterConnection instance to be used at server side where we have a
ConnectionRegistryEndpoint.IOExceptionCopyright © 2007–2020 The Apache Software Foundation. All rights reserved.