flume-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kowshik Prakasam <kows...@gmail.com>
Subject Re: Running multiple physical nodes on the same machine?
Date Tue, 20 Mar 2012 04:02:41 GMT
Hi Bharath,

One solution is to use the flume binary directly, instead of the daemon
script which doesn't seem
to allow multiple physical nodes to start up. Else you could try patching
the daemon script to
enalbe handling of multiple flume daemons.

Following are the commands to start-up multiple physical nodes using the
flume binary:

$ flume node_nowatch -n a
$ flume node_nowatch -n b


Thanks!
-Kowshik

On Mon, Mar 19, 2012 at 8:53 PM, Bharath Ravi <bharathravi1@gmail.com>wrote:

> Hi all,
>
> The FAQ here<https://cwiki.apache.org/FLUME/user-faq.html#UserFAQ-CanIruntwoinstancesoftheflumenodeonthesameunixmachine%253F>
says
> you can start multiple physical nodes on the same machine by simply
> specifying different names with the "-n" flag.
>
> This doesn't work for me.
>
> Here's a sample command I tried:
>
> # ./bin/flume-daemon.sh start node -n a
> starting node, logging to ...
> # ./bin/flume-daemon.sh start node -n b
> node running as process 24948. Stop it first.
>
> Am I missing something? What is the right way to do this?
>
> Thanks in advance,
> --
> Bharath Ravi
>
>
>
> On 19 March 2012 23:50, Bharath Ravi <bharathravi1@gmail.com> wrote:
>
>> On 19 March 2012 18:04, Jay Stricks <jay@wapolabs.com> wrote:
>>
>>> I couldn't get this to work either, so I started writing commands to a
>>> file and running: flume shell -c {master} -s {file}.
>>>
>>>
>> That's the workaround I've resorted too as well, for now. Is there a
>> better solution anyone knows of?
>>
>>
>>>
>>>
>>> On Mon, Mar 19, 2012 at 4:31 PM, Bharath Ravi <bharathravi1@gmail.com>wrote:
>>>
>>>> Hi all,
>>>>
>>>> The user guide<http://archive.cloudera.com/cdh/3/flume/UserGuide/#_using_the_flume_command_shell>
says
>>>> it is possible to pipe stdin into flume shell.
>>>> This does not work for me. Manually logging into the shell and running
>>>> the same commands works perfectly fine.
>>>> Any pointers as to what might be going wrong? Here's a command I tried
>>>> with its output:
>>>>
>>>>
>>>> # echo "connect fm-vm-010.cercs.int\nexec unconfig fm-vm-011.cercs.int"
>>>> | ./bin/flume shell
>>>> 2012-03-19 16:31:05,768 [main] INFO conf.FlumeConfiguration: Loading
>>>> configurations from conf
>>>> ==================================================
>>>> FlumeShell v0.9.4
>>>> Copyright (c) Cloudera 2010, All Rights Reserved
>>>> ==================================================
>>>> Type a command to execute (hint: many commands
>>>> only work when you are connected to a master node)
>>>>
>>>> You may connect to a master node by typing:
>>>>     connect host[:adminport=35873[:reportport=45678]]
>>>>
>>>> [flume (disconnected)] connect fm-vm-010.cercs.int\nexec unconfig
>>>> fm-vm-011.cercs.int
>>>> Failed to run command 'connect fm-vm-010.cercs.int\nexec unconfig
>>>> fm-vm-011.cercs.int' due to null
>>>> 2012-03-19 16:31:06,149 [main] ERROR util.FlumeShell: Failed to run
>>>> command 'connect fm-vm-010.cercs.int\nexec unconfig fm-vm-011.cercs.int
>>>> '
>>>>
>>>>
>>>>
>>>
>>
>

Mime
View raw message