Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Bigtop >> mail # user >> starting / stopping Big Top

Copy link to this message
Re: starting / stopping Big Top
Hi Glen,
Here is what I think.

1. The answer is a part of more general question of how to disable a linux
service from starting on bootup. For that, I personally found this link (
to be very helpful. If you use the method suggested there, and do something
sudo update-rc.d -f hadoop-hdfs-namenode remove

to disable the service on bootup.

You can verify it actually got disabled by doing something like:
sudo find /etc/ -name *hadoop-hdfs-datanode
You will see no references in /etc/rc*.d directories which are required for
starting up the service at the appropriate run level.

The above would stop the hadoop-hdfs-namenode from starting up
automatically. You will have to manually go and do something like:
sudo service hadoop-hdfs-namenode start

to start the service whenever you so desire.

If you later change your mind, and want to start the service on bootup
sudo update-rc.d hadoop-hdfs-namenode defaults

Once again,
sudo find /etc/ -name *hadoop-hdfs-namenode
would confirm if it actually got enabled. You will see links under

I don't know of a way on top of my head to do it for all of bigtop services
and the syntax doesn't seem to allow multiple services in one command. So,
you may have to do the disabling for the services one by one. A bash script
to do it for all services you care about would be handy.

It seems like rcconf tool that you suggested does the same thing as above,
so feel free to use that. I personally haven't used it though.

2. Yes, please go ahead document it and thanks for updating the wiki pages
in the past as well! Feel free to mention about all tools that you know so
users can pick and choose.

Please let us know if you have any further questions.


On Tue, Jan 29, 2013 at 7:33 AM, Glen Mazza <[EMAIL PROTECTED]> wrote:

> Hi, I followed the Wiki instructions (https://cwiki.apache.org/**
> confluence/display/BIGTOP/How+**to+install+Hadoop+**
> distribution+from+Bigtop+0.5.0<https://cwiki.apache.org/confluence/display/BIGTOP/How+to+install+Hadoop+distribution+from+Bigtop+0.5.0>
> **) to install Big Top on my Ubuntu 12.10-based laptop, all worked well.
>  Of course, Big Top creates several services mostly all prefixed hadoop-*
> that start automatically when I start/boot my laptop.  Two questions:
> 1.) Is there an easy way to start/stop the services that comprise Big Top
> as a unit, when I want to, instead of having them start up every time upon
> system reboot?  People just playing with/testing Big Top (like me) might
> not want those services to be always active on their laptops.
> 2.) I'd like to add instructions to the Wiki on how to disable the
> services.  Following this article: http://askubuntu.com/**
> questions/19320/whats-the-**recommended-way-to-enable-**disable-services<http://askubuntu.com/questions/19320/whats-the-recommended-way-to-enable-disable-services>,
> I've found rcconf (manpages.ubuntu.com/manpages/**natty/man8/rcconf.8.html<http://manpages.ubuntu.com/manpages/natty/man8/rcconf.8.html>)
> to be the fastest/easiest way to bulk enable/disable services at startup
> time.  Is there a better tool I should recommend instead?
> Thanks,
> Glen
> --
> Glen Mazza
> Talend Community Coders - coders.talend.com
> blog: www.jroller.com/gmazza