incubator-ambari-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
Subject svn commit: r1348298 - in /incubator/ambari/branches/ambari-186: CHANGES.txt README.txt
Date Sat, 09 Jun 2012 02:02:35 GMT
Author: reznor
Date: Sat Jun  9 02:02:35 2012
New Revision: 1348298

AMBARI-541. Update README for installing/running code off AMBARI-186. (Contributed by Vinod


Modified: incubator/ambari/branches/ambari-186/CHANGES.txt
--- incubator/ambari/branches/ambari-186/CHANGES.txt (original)
+++ incubator/ambari/branches/ambari-186/CHANGES.txt Sat Jun  9 02:02:35 2012
@@ -6,6 +6,8 @@ characters wide.
 Release 1.0.0 - unreleased
+  AMBARI-541. Update README for installing/running code off AMBARI-186. (Vinod K.V. via reznor)
   AMBARI-542. Rename HMC to Ambari in user-facing init.d scripts (Varun via Vikram)
   AMBARI-540. Naming cleanup required for management console. (Vikram via reznor)

Modified: incubator/ambari/branches/ambari-186/README.txt
--- incubator/ambari/branches/ambari-186/README.txt (original)
+++ incubator/ambari/branches/ambari-186/README.txt Sat Jun  9 02:02:35 2012
@@ -1,46 +1,112 @@
-How to use HMC:
-The rpm has a dependency on puppet. For now (until we have yum repo for HMC and puppet),
to install puppet on the install host, you will have to add the following repo.
+ - Ambari provides the GUI frontend that would help users install, manage and monitor the
Hadoop cluster.
+ - It uses CSS, Javascript, and HTML on client side. We use the existing proven
+open-source technologies in the backend e.g. PHP, Puppet for deploying and
+configuring the Hadoop services, Nagios & Ganglia for monitoring alerts and
+ metrics graphs respectively.
-edit file  /etc/yum.repos.d/puppet.repo as root
-to contain:
-baseurl =
+Getting Ambari
-yum install rpm-build
+The source can be checked out anonymously from SVN with this command:
-cd hmc; // The new directory structure
+  $ svn checkout ambari
-cd package/rpm
+How to build Ambari
-To install the rpm:
-sudo yum install php 5.1.6
-sudo yum install pdsh pdsh-rcmd-exec pdsh-rcmd-ssh
-sudo yum install httpd-devel mod_ssl
-sudo yum install puppet-2.7.9
-sudo yum install php-pdo-5.1.6
-sudo yum install php-pecl-json-1.2.1
-sudo yum install ruby-devel rubygems
-sudo rpm -Uvh
-sudo yum install mod_passenger
-sudo rpm -iv  build/rpmbuild/RPMS/noarch/hmc-0.0.1-2.noarch.rpm
+You need to build a AMBARI rpm from the source code that you checkout above:
-To Run:
+ $ yum install rpm-build
-Before starting, HMC requires java 32 bit and 64 bit to be available on the HMC host at /var/run/hmc/downloads/.
+ $ cd hmc; // The new directory structure
-sudo service hmc  start
+ $ cd package/rpm
+ $ ./
- http://HOSTNAME/hmc/html/index.php
+When it succeeds, you will find two RPMs inside the build directory package/rpm/build/rpmbuild/RPMS/noarch/:
- to get started.
+ 1) ambari-1.0.0-1.noarch.rpm
+ 2) ambari-agent-1.0.0-1.noarch.rpm
+Steps to run before you use Ambari for installing Hadoop on your cluster
+The node on which you will run the Ambari web-server will be henceforth referred
+to as Ambari master and the nodes which will be used to run Hadoop software will
+be called as cluster nodes.
+ 1) Set up password-less SSH for root to all your cluster nodes.
+    - Make sure you copy root's ssh public keys to all the cluster hosts.
+    - You will have to copy the ssh private key to the Ambari node for installing
+Hadoop on your cluster nodes from the UI.
  Make sure you copy root's ssh public keys to all the cluster hosts.
  You will have to copy the ssh private key to your desktop for later use in the UI.
+ 2) Install the above built ambari-agent-1.0.0-1.noarch.rpm on each of the nodes by running
+    $ sudo rpm -Uvh ambari-agent-1.0.0-1.noarch.rpm
+ 3) Stop ip-tables on your Ambari master:
+    $ sudo service iptables stop
+How to install Ambari
+Ambari has few external runtime dependencies, most important of which are
+ 1) puppet
+ 2) php
+ 3) ruby
+To simplify the installation of dependencies, you need to enable the following
+yum repos. For that, you can simply install the corresponding RPMs.
+ 1) EPEL repo
+ 2) Puppet passenger
+ 3) Hadoop repo
+So, you should do the following at the command line:
+ $ sudo rpm -Uvh[i386|x86_64]/epel-release-5-4.noarch.rpm

+ $ sudo rpm -Uvh

+ $ sudo rpm -Uvh
+Now, to install ambari-rpms
+ $ sudo yum install php-pecl-json (version 1.2.1 compatible with php-5.1 or php-5.2)
+ $ sudo rpm -iv ambari-1.0.0-1.noarch.rpm 
+Hadoop required JDK, so you can download it from the Oracle website
+ . Ambari needs the following
+two files to be available on the Ambari master node under /var/run/hmc/downloads/
+  1) jdk-6u26-linux-x64.bin
+  2) jdk-6u26-linux-i586.bin
+How to run Ambari
+ $ sudo service ambari start
+You have installed ambari by now. You can start deploying Apache Hadoop and
+ecosystem components on your cluster nodes by visiting:
+ http://AMBARIMASTER/hmc/html/index.php

View raw message