Skip to content


Subversion checkout URL

You can clone with
Download ZIP
The AMI takes a set of input parameters via the EC2 user-data to install, RAID, ring, and launch a DataStax Enterprise/Community cluster.
Python HTML PHP Shell

Update ami_ids.json and change format

With more attributes than just hvm vs pv, the hierarchical scheme
doesn't really hold up anymore. Top-level keys are now regions and
inside that is just a list of objects representing ami's which can now
have arbitrary-ish keys if we need to add metadata.
latest commit 0114ef483e
@mlococo mlococo authored


DataStax's Amazon Machine Image is the quickest way to get a DataStax Community or DataStax Enterprise cluster up and running on EC2.


Use cassandralauncher as found and documented here:

This will ensure all options are processed correctly and easily.


Basic AMI Switches:

--clustername <name>
    The name of the Cassandra cluster

--totalnodes <#>
    Cluster size

--version [ community | enterprise ]
    Installs either DataStax Enterprise or
    DataStax Community Edition

    Binds the rpc_address to the private IP
    address of the instance
    Default: false, uses

DataStax Enterprise Specific:

--username <user>
    The username provided during DSE registration
    --password is REQUIRED in order to use this option
    REQUIRED for a DSE installation

--password <pass>
    The password provided during DSE registration
    --username is REQUIRED in order to use this option
    REQUIRED for a DSE installation

--analyticsnodes <#>
    Number of analytics nodes that run with Spark
    Note: Uses Hadoop in versions earlier than DSE 4.5
    Default: 0

--searchnodes <#>
    Number of search nodes that run with Solr
    Default: 0

    Force Hadoop over Spark on analytics nodes
    Default: false, uses Spark on 4.5+


--release <release_version>
    Allows for the installation of a previous DSE version
    Example: 1.0.2
    Default: Ignored

--cfsreplicationfactor <#>
    The CFS replication factor
    Note: --cfsreplicationfactor must be <= --analyticsnodes
    Default: 1

--opscenter no
    Disables the installation of OpsCenter on the cluster
    Default: yes

--reflector <url>
    Allows you to use your own reflector

--repository <repository>
    Allows you to set a custom repository to pull configuration files from
    Default: none, falls back to repository used to create the AMI

Ports Needed



Visit for full installation instructions.


To stop the service, simply run

sudo service <cassandra | dse> stop

To start the service again, simply run

sudo service <cassandra | dse> start

Implementation details

See FILES.txt for a description of how the scripts here configure the AMI.


  1. Backup all the data on all your nodes using the snapshot utility. This provides you with the easiest way to revert any unwanted changes or incompatibilities that may arise. See the DataStax documentation for more information.
  2. On each of your Cassandra nodes, run sudo apt-get install [ cassandra | apache-cassandra1 | dse-full ], depending on which version you were currently on and want to upgrade to.
    • cassandra upgrades to the latest in 0.8.x release.
    • apache-cassandra upgrades to the latest in the 1.0.x release.
    • dse-full upgrades to the latest DataStax Enterprise release.
    • If you are trying to upgrade across major versions, make sure to read NEWS.txt on the newer packages and consult DataStax documentation for full details for upgrading packaged releases. Typically a new repository must be added followed by a sudo apt-get update.
  3. Account for New and Changed Parameters in cassandra.yaml. If the default Cassandra configuration file has changed, you will find backups of it in the conf directory. You can use that to compare the two configurations and make appropriate changes.
  4. Make sure any client drivers – such as Hector or Pycassa clients – are compatible with your current version.
  5. Run nodetool drain to flush the commit log and then restart each Cassandra node, one at a time, monitoring the log files for any issues.
  6. After upgrading and restarting all Cassandra nodes, restart client applications.
  7. [Upgrading from 0.8 to 1.0] After upgrading, run nodetool scrub against each node before running repair, moving nodes, or adding new ones.

Branching details

Feel free to fork off this project and offer any suggestions that you find along the way.

Also, if you're interested in the whole process: read up on the saving process here:

Something went wrong with that request. Please try again.