Author: ctubbsii
Date: Wed Jun 5 16:55:47 2013
New Revision: 1489956
URL: http://svn.apache.org/r1489956
Log:
ACCUMULO-935 Apply README patch from David Medinets, with my modifications for
clarity and to expand on the behavior of DEV_ACCUMULO_HOME
Modified:
accumulo/trunk/README
Modified: accumulo/trunk/README
URL:
http://svn.apache.org/viewvc/accumulo/trunk/README?rev=1489956&r1=1489955&r2=1489956&view=diff
==============================================================================
--- accumulo/trunk/README (original)
+++ accumulo/trunk/README Wed Jun 5 16:55:47 2013
@@ -15,30 +15,43 @@ ready to go on x86 GNU/Linux: there is n
However, if you only have source code, or you wish to make changes, you need to
have maven configured to get Accumulo prerequisites from repositories. See
-the pom.xml file for the necessary components. Activate the 'docs' profile to
build
-the Accumulo developer and user manual.
+the pom.xml file for the necessary components.
-Run "mvn package -P assemble" to build a distribution, or run
-"mvn package -P assemble,docs" to also build the documentation. By default,
-Accumulo compiles against Hadoop 1.0.4. To compile against a different version
-that is compatible with Hadoop 1.0, specify hadoop.version on the command line,
-e.g. "-Dhadoop.version=0.20.205.0" or "-Dhadoop.version=1.1.0". To compile
-against Hadoop 2.0, specify "-Dhadoop.profile=2.0". By default this uses
+You can build an Accumulo binary distribution, which is created in the
+assemble/target directory, using the following command. Note that maven 3
+is required starting with Accumulo v1.5.0. By default, Accumulo compiles
+against Hadoop 1.0.4.
+
+ mvn package -P assemble
+
+To compile against a different version that is compatible with Hadoop 1.0,
+specify hadoop.version on the command line,
+e.g. "-Dhadoop.version=0.20.205.0" or "-Dhadoop.version=1.1.0".
+
+To compile against Hadoop 2.0, specify "-Dhadoop.profile=2.0". By default
this uses
2.0.4-alpha. To compile against a different 2.0-compatible version, specify
the profile and version, e.g. "-Dhadoop.profile=2.0 -Dhadoop.version=0.23.5".
If you are running on another Unix-like operating system (OSX, etc) then
you may wish to build the native libraries. They are not strictly necessary
-but having them available suppresses a runtime warning:
+but having them available suppresses a runtime warning and enables Accumulo
+to run faster:
$ ( cd ./server/src/main/c++ ; make )
-If you want to build the debian release, use the command "mvn install -Pdeb"
to
+If you want to build the debian release, use the command "mvn package -Pdeb"
to
generate the .deb files in the target/ directory. Please follow the steps at
https://cwiki.apache.org/BIGTOP/how-to-install-hadoop-distribution-from-bigtop.html
to add bigtop to your debian sources list. This will make it substantially
easier to install.
+Building Documentation
+
+Use the following command to build the User Manual
(docs/target/accumulo_user_manual.pdf)
+and the configuration HTML page (docs/target/config.html)
+
+ mvn package -P docs -DskipTests
+
******************************************************************************
2. Deployment
@@ -49,6 +62,39 @@ tar xzf accumulo-1.6.0-bin.tar.gz).
If you are using the RPM, install the RPM on every machine that will run
accumulo.
+Another option is to package Accumulo directly to a working directory. For
example,
+
+ mvn package -DskipTests -DDEV_ACCUMULO_HOME=/var/tmp
+
+The above command would create a directory with a name similar to
+/var/tmp/accumulo-1.6.0-dev/accumulo-1.6.0/, containing all the contents
+that are normally contained in accumulo-1.6.0-bin.tar.gz, but already unpacked.
+If the DEV_ACCUMULO_HOME parameter is not specified, this directory would
+normally be created in assemble/target, but that is subject to deletion by
+the 'mvn clean' command. Specifying an external directory would not be subject
+to 'mvn clean'. When executed more than once, newer files overwrite older
files,
+and files a user adds (such as configuration files in conf/) will be left
alone.
+
+If HDFS and Zookeeper are running, you can run Accumulo directly from this
+working directory. See the 'Running Apache Accumulo' section later in this
document.
+
+You can avoid specifying the working directory each time you compile by adding
+a profile to maven's settings.xml file. Below is an example of
$HOME/.m2/settings.xml
+
+ <settings xmlns="http://maven.apache.org/SETTINGS/1.0.0";
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance";
xsi:schemaLocation="http://maven.apache.org/SETTINGS/1.0.0
http://maven.apache.org/xsd/settings-1.0.0.xsd";>
+ <profiles>
+ <profile>
+ <id>inject-accumulo-home</id>
+ <properties>
+ <DEV_ACCUMULO_HOME>/var/tmp</DEV_ACCUMULO_HOME>
+ </properties>
+ </profile>
+ </profiles>
+ <activeProfiles>
+ <activeProfile>inject-accumulo-home</activeProfile>
+ </activeProfiles>
+ </settings>
+
******************************************************************************
3. Upgrading from 1.4 to 1.5
@@ -64,10 +110,10 @@ accumulo.
4. Configuring
Apache Accumulo has two prerequisites, hadoop and zookeeper. Zookeeper must be
-at least version 3.3.0. Both of these must be installed and configured.
-Zookeeper normally only allows for 10 connections from one computer. On a
-single-host install, this number is a little too low. Add the following to
the
-$ZOOKEEPER_HOME/conf/zoo.cfg file:
+at least version 3.3.0. Both of these must be installed and configured. Some
+versions of Zookeeper may only allow 10 connections from one computer by
default.
+On a single-host install, this number is a little too low. Add the following to
+the $ZOOKEEPER_HOME/conf/zoo.cfg file:
maxClientCnxns=100
