You can have multiple source-sink pairs on the same physical machine.  Each
of those has to have its own flume process running and its own
configurations, though. So you give each a unique name, and then execute
the {$FLUME_HOME}/bin/flume node -n {node_name} for each, or adjust the
initialization script you're using.

2012/3/5 lulynn_2008 <lulynn_2008@163.com>

> Yes, there is already a node before we start. Do you mean we just can
> start 1 agent node in a host?
>
>
>
> At 2012-03-05 20:02:53,"alo alt" <wget.null@googlemail.com> wrote:
> >Hi,
> >
> >1. as the message says - address in use. Check with ps if more as one process is running at one time. Stop the node, do ps and kill all processes with flume inside (ps waux|grep flume, if the process are runs as flume you can try killall -9 -u flume)
> >2. as in the mail before, 0.94 does not work with hadoop-1.0.0
> >3. the flume master isn't available, or the port is blocked or network issue or something like that
> >
> >- Alex
> >
> >
> >--
> >Alexander Lorenz
> >http://mapredit.blogspot.com
>
> >
> >On Mar 5, 2012, at 12:40 PM, lulynn_2008 wrote:
> >
> >>  Hi,
> >> -  Every time I start a agent node, the following WARN appears. Is there any way to avoid this?
> >> -  Besides, I found this "Flume is using Hadoop core 1.0.0 which does not support Security / Authentication: null". Is this mean flume-0.9.4 can not work with hadoop-1.0.0?
> >> -  Is there any problem with this "No active master RPC connection"?
> >>
> >> 2012-03-05 03:02:04,483 [main] WARN util.InternalHttpServer: Caught exception during HTTP server start.
> >> java.net.BindException: Address already in use
> >>         at sun.nio.ch.Net.bind(Native Method)
> >>         at sun.nio.ch.ServerSocketChannelImpl.bind(ServerSocketChannelImpl.java:156)
> >>         at sun.nio.ch.ServerSocketAdaptor.bind(ServerSocketAdaptor.java:71)
> >>         at org.m ortbay.jetty.nio.SelectChannelConnector.open(SelectChannelConnector.java:216)
> >>         at org.mortbay.jetty.nio.SelectChannelConnector.doStart(SelectChannelConnector.java:315)
> >>         at org.mortbay.component.AbstractLifeCycle.start(AbstractLifeCycle.java:50)
> >>         at org.mortbay.jetty.Server.doStart(Server.java:235)
> >>         at org.mortbay.component.AbstractLifeCycle.start(AbstractLifeCycle.java:50)
> >>         at com.cloudera.util.InternalHttpServer.start(InternalHttpServer.java:188)
> >>         at com.cloudera.flume.agent.FlumeNode.start(FlumeNode.java:275)
> >>         at com.cloudera.flume.agent.FlumeNode.setup(FlumeNode.java:504)
> >>         at com.cloudera.flume.agent.FlumeNode.main(FlumeNode.j ava:665)
> >> 2012-03-05 03:02:04,485 [main] ERROR agent.FlumeNode: Unexpected exception/error thrown! Unable to start HTTP server
> >> com.cloudera.util.InternalHttpServer$InternalHttpServerException: Unable to start HTTP server
> >>         at com.cloudera.util.InternalHttpServer.start(InternalHttpServer.java:194)
> >>         at com.cloudera.flume.agent.FlumeNode.start(FlumeNode.java:275)
> >>         at com.cloudera.flume.agent.FlumeNode.setup(FlumeNode.java:504)
> >>         at com.cloudera.flume.agent.FlumeNode.main(FlumeNode.java:665)
> >> Caused by: java.net.BindException: Address already in use
> >>         at sun.nio.ch.Net.bind(Native Method)
> >>         at sun.nio.ch.ServerSocketChannelImpl.bind(ServerSocketChannelImpl.java:156)
> >>         at sun. nio.ch.ServerSocketAdaptor.bind(ServerSocketAdaptor.java:71)
> >>         at org.mortbay.jetty.nio.SelectChannelConnector.open(SelectChannelConnector.java:216)
> >>         at org.mortbay.jetty.nio.SelectChannelConnector.doStart(SelectChannelConnector.java:315)
> >>         at org.mortbay.component.AbstractLifeCycle.start(AbstractLifeCycle.java:50)
> >>         at org.mortbay.jetty.Server.doStart(Server.java:235)
> >>         at org.mortbay.component.AbstractLifeCycle.start(AbstractLifeCycle.java:50)
> >>         at com.cloudera.util.InternalHttpServer.start(InternalHttpServer.java:188)
> >>         ... 3 more
> >> 2012-03-05 03:02:04,527 [Heartbeat] INFO agent.MultiMasterRPC: No active master RPC connection
> >> 2012-03-05 03:02:04,568 [Heartbeat] INFO agent. ThriftMasterRPC: Connected to master at svltest150.svl.ibm.com:35872
> >> 2012-03-05 03:02:04,711 [main] INFO agent.LogicalNodeManager: creating new logical node agent
> >> 2012-03-05 03:02:04,879 [main] WARN agent.FlumeNode: Flume is using Hadoop core 1.0.0 which does not support Security / Authentication: null
> >>
> >>
> >
>
>
>
>