Hi,

sent via my mobile device

On Apr 1, 2012, at 4:44 PM, Mohit Anchlia <[email protected]> wrote:

inline

> Thanks for the post.
>  
> I have some questions:
>  
> 1. Is writing to the SequenceFile better? I thought sequenceFile is generally 
> good for binary data.

Yes. 

>  
> 2. Generally is one agent sufficient for one client? If not then idea is to 
> just start multiple agents?

Yes, put them into the same channel

>  
> 3. How to deal with failures of agents or clients?

Use a couple of agents, and in front if them an loadbalancer as example. 

best,
- Alex

> 
> On Sun, Apr 1, 2012 at 2:39 AM, alo alt <[email protected]> wrote:
> I wrote a couple of weeks ago a blog post:
> http://mapredit.blogspot.de/2012/03/flumeng-evolution.html
> 
> here u found all what you need.
> 
> - Alex
> 
> --
> Alexander Lorenz
> http://mapredit.blogspot.com
> 
> On Apr 1, 2012, at 8:26 AM, shekhar sharma wrote:
> 
> > Hello Mohit,
> > Create a property file "foo.properties" and then run the following command 
> > to start your agent
> > ./flume-ng  node --conf ../conf/ -f  ../conf/foo.properties -n foo
> >
> > and then you can run ur avro client to send some sample data which will be 
> > collected by the foo's source an it agent i.e. avro source which then it 
> > will send to HDFS.
> >
> > flume-ng avro-client --conf ../conf -H localhost -p 41414 -F /etc/passwd
> >
> >
> > you need to define the host and the port for foo.source in your 
> > foo.properties file. Hope that helps.
> >
> > Regards,
> > Som
> >
> >
> > On Sat, Mar 31, 2012 at 5:27 AM, Mohit Anchlia <[email protected]> 
> > wrote:
> > I am trying to test how flume ng works so I create a conf and tried to run:
> >
> >  foo.sources = avroSrc
> > foo.channels = memoryChannel
> > foo.sinks = hdfsSink
> > # For each one of the sources, the type is defined
> > foo.sources.avroSrc.type = avro
> > # The channel can be defined as follows.
> > foo.sources.avroSrc.channels = memoryChannel
> > # Each sink's type must be defined
> > foo.sinks.hdfsSink.type = hdfs
> > foo.sinks.hdfsSink.path = hdfs://dsdb1:9000/flume
> > #Specify the channel the sink should use
> > foo.sinks.hdfsSink.channel = memoryChannel
> > # Each channel's type is defined.
> > foo.channels.memoryChannel.type = memory
> > # Other config values specific to each type of channel(sink or source)
> > # can be defined as well
> > # In this case, it specifies the capacity of the memory channel
> > foo.channels.memoryChannel.capacity = 100
> >
> > But java process is not running:
> >
> > [root@dslg1 flume-1.2.0-incubating-SNAPSHOT]# ./bin/flume-ng  node --conf 
> > ../conf/flume-conf.properties
> > + exec /usr/java/jdk1.6.0_24/bin/java -Xmx20m -cp 
> > '../conf/flume-conf.properties:/root/.mohit/flume-1.2.0-incubating-SNAPSHOT/lib/*'
> >  -Djava.library.path= org.apache.flume.node.Application
> > log4j:WARN No appenders could be found for logger 
> > (org.apache.flume.node.Application).
> > log4j:WARN Please initialize the log4j system properly.
> > log4j:WARN See http://logging.apache.org/log4j/1.2/faq.html#noconfig for 
> > more info.
> > --
> > Can you please suggest what I might be doing wrong?
> >
> >
> >
> 
> 

Reply via email to