This is the documentation for CDH 5.0.x. Documentation for other versions is available at Cloudera Documentation.

Llama Installation

Llama is a system that mediates resource management between Cloudera Impala and Hadoop YARN. Llama enables Impala to reserve, use, and release resource allocations in a Hadoop cluster. Llama is only required if resource management is enabled in Impala.
  Note:
  • Impala does not currently support Llama with Kerberos.
  • At this point Llama has been tested only in a Cloudera Manager deployment. For information on using Cloudera Manager to configure Llama and Impala, see Installing Impala with Cloudera Manager.
See the following sections for information and instructions:

Packaging

The packaging options for installing Llama are:

  • A RPM package for the Llama ApplicationMaster(llama-master)
  • A Debian package for the Llama ApplicationMaster(llama-master)

Installing Llama

Llama is distributed in two packages:
  • llama- the binaries and configuration files
  • llama-master - the service script that you use to run Llama
Installing the llama-master package installs the llama package and the dependencies needed to run Llama, creating a llama service configured to start Llama at system startup time.
  Important:

If you have not already done so, install Cloudera's yum, zypper/YaST or apt repository before using the following commands to install Llama. For instructions, see CDH 5 Installation.

To install Llama on a Red Hat system:

$ sudo yum install llama-master

To install Llama on an Ubuntu or other Debian system:

$ sudo apt-get install llama-master

To install Llama on a SLES system:

$ sudo zypper install llama-master

Configuring Llama

  Important:

Llama works with YARN only; it does not work with MRv1.

When you install Llama from an RPM or Debian package, Llama server creates all configuration, documentation, and runtime files in the standard Linux directories, as follows.

Type of File

Where Installed

Binaries

/usr/lib/llama/

Configuration

/etc/llama/conf/

Logs

/var/log/llama/

PID file

/var/run/llama/

Llama uses the YARN configuration to interact with Hadoop. The Llama configuration file, /etc/llama/conf/llama-site.xml, contains all default values after installation. You do not need to change these to get Llama up and running.

Starting, Stopping, and Using Llama

To start Llama

After you have completed all the required configuration steps, you can start Llama:
$ sudo service llama start
If you see the message LlamaAMServer - Llama started! in the llama.log log file, the system has started successfully.

To stop Llama

Stop Llama as follows
$ sudo service llama stop
Page generated September 3, 2015.