You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@ambari.apache.org by om...@apache.org on 2011/11/15 18:18:16 UTC
svn commit: r1202315 - in /incubator/ambari/trunk/src/site/apt: cli.apt
scenarios.apt
Author: omalley
Date: Tue Nov 15 17:18:15 2011
New Revision: 1202315
URL: http://svn.apache.org/viewvc?rev=1202315&view=rev
Log:
AMBARI-111. Minor clean up of site documentation (omalley)
Modified:
incubator/ambari/trunk/src/site/apt/cli.apt
incubator/ambari/trunk/src/site/apt/scenarios.apt
Modified: incubator/ambari/trunk/src/site/apt/cli.apt
URL: http://svn.apache.org/viewvc/incubator/ambari/trunk/src/site/apt/cli.apt?rev=1202315&r1=1202314&r2=1202315&view=diff
==============================================================================
--- incubator/ambari/trunk/src/site/apt/cli.apt (original)
+++ incubator/ambari/trunk/src/site/apt/cli.apt Tue Nov 15 17:18:15 2011
@@ -71,6 +71,12 @@ Command Line Interface
* ambari add-user
+ * <<Server commands>>
+
+ * ambari controller
+
+ * ambari agent
+
* <<Generic commands>>
* ambari help
@@ -96,10 +102,17 @@ Cluster Commands
* Name of the user defined stack defining cluster configuration
- * <<-stack_revision>> [ STACK REVISION ]
+ * <<-stack_revision>> [ INTEGER ]
* Revision of the stack. If not specified latest one is used.
+ * <<-stack-file>> [ FILENAME]
+
+ * Update the stack with the same name as the cluster using the
+ given file and use the updated stack for creating the new
+ cluster. This option is mutually exclusive with <<-stack>> and
+ <<-stack_revision>>
+
* <<-nodes>> [node_range_exp1, node_range_exp2] <<[REQUIRED]>>
* Specify range of nodes associated with the cluster. One or
@@ -118,10 +131,10 @@ Cluster Commands
* Default is INACTIVE
- * <<-services>> [component-1, component-2, component-3]
+ * <<-components>> [component-1, component-2, component-3]
* By default all the components will be activated upon cluster
- activation. If user specifies specific ones then only they
+ activation. If user specifies specific ones then only those
will be activated upon cluster activation
* <<-role>> rolename=[node_range_exp1, node_range_exp2, â¦]
@@ -138,7 +151,8 @@ Cluster Commands
* <<-wait>>
- * Optionally wait for cluster to get to goal state
+ * Optionally wait for cluster to get to goal state. The progress of
+ activating the cluster will be printed to the user.
[]
@@ -155,15 +169,23 @@ Cluster Commands
* <<-stack>> [ STACK NAME ]
- * <<-stack_revision>> [ STACK REVISION ]
+ * <<-stack_revision>> [ INTEGER ]
* Revision of the stack. If not specified latest one is used.
+ * <<-stack-file>> [ FILENAME]
+
+ * Update the stack with the same name as the cluster using the
+ given file and use the updated stack in the updated
+ cluster. This option is mutually exclusive with <<-stack>> and
+ <<-stack_revision>>
+
* <<-goalstate>> [ACTIVE/INACTIVE/ATTIC]
- * <<-services>> [component, component, component]
+ * <<-components>> [component, component, component]
- * Specify the list of desired active components
+ * Specify the list of desired active components that will be run
+ when the cluster is active.
* <<-nodes>> [node_range_exp1, node_range_exp2]
@@ -174,7 +196,8 @@ Cluster Commands
* <<-role>> rolename=[node_exp1, node_exp2, â¦] <<[REPEATABLE]>>
* Change the nodes associated with each role. The nodes will be
- re-deployed and re-configured appropriately.
+ re-deployed and re-configured appropriately. If nodes are not assigned
+ to a role, Ambari will automatically assign them.
* <<-dry-run>>
@@ -184,7 +207,8 @@ Cluster Commands
* <<-wait>>
- * Wait for cluster to get to desired goal state
+ * Wait for cluster to get to desired goal state. The progress of
+ activating the cluster will be printed to the user.
[]
@@ -231,7 +255,8 @@ Cluster Commands
* <<-wait>>
- * Optionally wait for cluster to be deleted
+ * Optionally wait for cluster to be deleted. The progress of deleting
+ the cluster will be displayed to the user.
[]
@@ -453,9 +478,9 @@ Configuration Commands
* Set the password for the Ambari agent to authenticate to the
controller.
- * <<-controller>> [MACHINE NAME]
+ * <<-controller>> [CONTROLLER URL]
- * Set the machine name for the Ambari controller.
+ * Set the URL for the Ambari controller REST API.
* <<ambari add-user>>
@@ -475,14 +500,36 @@ Configuration Commands
* The user should provide the given password when authenticating.
+Server Commands
+
+ * <<ambari controller>>
+
+ * Start the Ambari controller on this machine
+
+ * <<Options:>>
+
+ * <<-autostart>> [BOOLEAN]
+
+ * In addition to starting the server, update the system to
+ (not) restart the controller when the system reboots.
+
+ * <<ambari agent>>
+
+ * Start the Ambari agent on this machine
+
+ * <<Options:>>
+
+ * <<-autostart>> [BOOLEAN]
+
+ * In addition to starting the server, update the system to
+ (not) restart the agent when the system reboots.
+
Generic Commands
* <<ambari help>>
* Provides the CLI help
- []
-
* <<ambari version>>
* Ambari CLI version
Modified: incubator/ambari/trunk/src/site/apt/scenarios.apt
URL: http://svn.apache.org/viewvc/incubator/ambari/trunk/src/site/apt/scenarios.apt?rev=1202315&r1=1202314&r2=1202315&view=diff
==============================================================================
--- incubator/ambari/trunk/src/site/apt/scenarios.apt (original)
+++ incubator/ambari/trunk/src/site/apt/scenarios.apt Tue Nov 15 17:18:15 2011
@@ -25,18 +25,21 @@ Ambari Example Scenarios
$ rpm -i ambari-0.1.0-1.rpm
$ ambari configure -agent-password my.pw
$ ambari add-user -user sue -kerberos
-$ ambari controller
+$ ambari -autostart true controller
-------
- On slaves using pdsh:
+ If you have password-less ssh and pdsh installed, you can install the agents
+ directly on node00 to node99 using:
-------
-$ rpm -i ambari-0.1.0-1.rpm
-$ ambari configure -controller controller.my.domain.com -agent-password my.pw
-$ ambari agent
+$ pdsh -w node00-99 \
+ 'rpm -i ambari-0.1.0-1.rpm; \
+ ambari configure -controller controller.my.domain.com \
+ -agent-password my.pw ; \
+ ambari -autostart true agent'
-------
-* Creating a cluster using CLI
+* Creating a simple cluster using CLI
Create file cluster1.json with the specific configuration
alterations for cluster1.
@@ -48,32 +51,17 @@ $ ambari agent
{"@name": "ambari",
"property": [
{"@name": "data.dirs", "@value": "/data/*"},
- {"@name": "keytab.dir", "@value": " /etc/security/keytab"},
- {"@name": "user.realm", "@value": "MY.DOMAIN.COM"}]},
- {"@name": "hadoop-env",
- "property": [
- {"@name": "HADOOP_CLIENT_OPTS",
- "@value": "-Xmx256m ${HADOOP_CLIENT_OPTS}"}]}]},
- "components": [
- {"@name":"hdfs",
- "configuration": {
- "category": [
- {"@name": "hadoop-env",
- property: [
- {"@name": "HADOOP_NAMENODE_OPTS",
- "@value": "-Xmx512m -Dsecurity.audit.logger=INFO,DRFAS
- -Dhdfs.audit.logger=INFO,DRFAAUDIT"}]}]}}]}
+ {"@name": "user.realm", "@value": "MY.DOMAIN.COM"}]}]}}
-------
- Run the commands using the CLI and Kerberos authentication. The first command
- creates the stack and the second creates a cluster based on that stack using
- the machines host00-99.
+ Run the following commands using the CLI.
+ The first command creates the stack and the second
+ creates a cluster based on that stack using the machines host00-99.
-------
-$ ambari stack create -name cluster1 -file cluster1.json
-$ ambari cluster create -name cluster1 -stack cluster1 -nodes host00-99 \
- -goalstate active -role namenode=host00 -role jobtracker=host01 \
- -role client=host98-99 -wait
+$ ambari cluster create -name cluster1 -stack-file cluster1.json \
+ -nodes host00-99 -goalstate active -role namenode=host00 \
+ -role jobtracker=host01 -role client=host98-99 -wait
-------
* Upgrading a cluster using CLI
@@ -91,8 +79,43 @@ $ ambari cluster create -name cluster1 -
for the command to complete.
-------
-$ ambari stack update -name cluster1 -file cluster1-update.json
-$ ambari cluster update -name cluster1 -stack cluster1 -wait
+$ ambari cluster update -name cluster1 -stack-file cluster1-update.json -wait
+-------
+
+* Creating a more complicated cluster using CLI
+
+ If you want to bump up the memory for the clients' JVM to 256MB and the
+ NameNode to 512MB, you can define a cluster like the below:
+
+------
+{"@parentName":"hadoop-security",
+ "configuration": {
+ "category": [
+ {"@name": "ambari",
+ "property": [
+ {"@name": "data.dirs", "@value": "/data/*"},
+ {"@name": "keytab.dir", "@value": " /etc/security/keytab"},
+ {"@name": "user.realm", "@value": "MY.DOMAIN.COM"}]},
+ {"@name": "hadoop-env",
+ "property": [
+ {"@name": "HADOOP_CLIENT_OPTS", "-Xmx256m"}]}]},
+ "components": [
+ {"@name":"hdfs",
+ "configuration": {
+ "category": [
+ {"@name": "hadoop-env",
+ property: [
+ {"@name": "HADOOP_NAMENODE_OPTS",
+ "@value": "-Xmx512m -Dsecurity.audit.logger=INFO,DRFAS
+ -Dhdfs.audit.logger=INFO,DRFAAUDIT"}]}]}}]}
+------
+
+ And run the command to create the new cluster:
+
+-------
+$ ambari cluster create -name cluster2 -stack-file cluster2.json \
+ -nodes host00-99 -goalstate active -role namenode=host00 \
+ -role jobtracker=host01 -role client=host98-99 -wait
-------
* Creating a cluster using REST