Install DataStax Enterprise (DSE) 6.8 with a binary tarball

You can use a binary tarball to install DataStax Enterprise (DSE) 6.8 on bare metal or a virtual machine (VM) running a supported Linux platform.

For new deployments, DataStax recommends the latest patch release. If you are scaling an existing deployment, install the same patch release as the existing nodes. For a summary of features and changes in each release, see the DSE 6.8 release notes.

When installed from a tarball, DSE runs as a standalone process, with or without root permissions.

When you install DSE, it automatically creates a cassandra user that runs the DSE process. For security reasons, don’t use the default cassandra user in production. After installation, create a new superuser, and then disable the cassandra user.

For other installation options, see Choose an installation method.

Prerequisites

  • Prepare an environment where you want to install DSE, including sufficient hardware and a supported platform.

  • For RHEL-compatible platforms, enable Extra Packages for Enterprise Linux (EPEL).

  • Install a supported Java runtime: OpenJDK (recommended) or Oracle Java SE (JDK).

    Java 11 can be used for core (transactional) workloads in DSE 6.8.22 and later. Java 8 is required for advanced workloads in all 6.8 patch releases.

    If you install both Java 8 and Java 11, set your $JAVA_HOME environment variable to Java 11. Otherwise, set $JAVA_HOME to Java 8.

  • To run the CQL shell (cqlsh), install a supported Python version. DataStax recommends 3.8 to 3.11.

  • For production installations and simulated production test environments, review the recommended settings.

    Some settings can be applied before installing DSE. For settings that require a running DSE instance, plan to apply them after the installation.

Download and extract the tarball

  1. Install the libaio package that matches your environment:

    RHEL
    sudo yum install libaio
    Debian
    sudo apt-get install libaio1
    Ubuntu 24.04 and later

    On Ubuntu 24.04 and later, the libaio library was renamed to libaio1t64. DSE requires a symlink from the old library name to the new one to detect it properly. Without this link, DSE issues a warning about the missing library and performance is significantly degraded.

    		 
    sudo apt-get install libaio1t64
    sudo ln -s /lib/x86_64-linux-gnu/libaio.so.1t64 /lib/x86_64-linux-gnu/libaio.so.1

  2. Find DSE 6.8 on IBM Fix Central:

    1. Go to Fix Central.

    2. In the Product selector field, begin typing IBM DataStax Enterprise, and then select that option from the menu.

    3. In the Release field, select the version that you want to install.

    4. Click Continue.

    5. On the Identify fixes page, click Continue to use the default Browse for fixes option.

    6. On the Select fixes page, select the fix pack (DSE version) you want to install, and then click Continue.

      Depending on the product and version, more fix packs might be available if you set the Platform filter to All, and then click Submit.

    7. If prompted, sign in with your IBMid.

      An IBMid account with MFA enabled is required. If you don’t have one, create an IBMid account. If your organization uses Enterprise Federation (EF) for authentication with corporate credentials, see the IBMid EF documentation.

    8. On the Download options page, select Download using your browser (HTTPS), and then click Continue.

    9. Review the terms and conditions, and then click I agree to activate the download links.

  3. Click the DSE tarball link to download the file, for example, dse-6.8.60-bin.tar.gz.

  4. Extract the tarball file into the directory where you want to install DSE:

    sudo tar -xzvf dse-6.8.60-bin.tar.gz -C INSTALLATION_DIRECTORY

    The files expand into a DSE_DIRECTORY that contains the version number of the release, such as dse-6.8.60.

Configure data and log directories

  1. Use either the default DSE data and log directory locations, or define your own custom directory locations:

    Default directory locations

    DSE stores its runtime data and logs in default locations unless otherwise configured in cassandra.yaml. Make sure these directories exist and have the correct ownership before starting DSE.

    Run the following commands to create and change ownership for the default DSE data and log directories:

    sudo mkdir -p /var/lib/cassandra /var/log/cassandra
    sudo chown -R cassandra:cassandra /var/lib/cassandra /var/log/cassandra

    If you plan to run DSE as a different user, replace cassandra:cassandra with that user and group.

    Custom directory locations

    To store DSE data and logs in custom locations, create and change ownership for your custom DSE data and log directories, replacing CUSTOM_DIRECTORY with the absolute path to your custom directory locations. For example:

    sudo mkdir -p \
  CUSTOM_DIRECTORY/dse-data/commitlog \
  CUSTOM_DIRECTORY/dse-data/saved_caches \
  CUSTOM_DIRECTORY/dse-data/hints \
  CUSTOM_DIRECTORY/cdc_raw
    sudo chown -R cassandra:cassandra \
  CUSTOM_DIRECTORY/dse-data \
  CUSTOM_DIRECTORY/cdc_raw

    If you plan to run DSE as a different user, replace cassandra:cassandra with that user and group.

    Then, update the following settings in the cassandra.yaml file to point to your custom directory locations. For tarball installations, the default location of the cassandra.yaml file is <installation_location>/resources/cassandra/conf/cassandra.yaml.

    data_file_directories:
  - CUSTOM_DIRECTORY/dse-data
commitlog_directory: CUSTOM_DIRECTORY/dse-data/commitlog
saved_caches_directory: CUSTOM_DIRECTORY/dse-data/saved_caches
hints_directory: CUSTOM_DIRECTORY/dse-data/hints
cdc_raw_directory: CUSTOM_DIRECTORY/cdc_raw

  2. To store logs and data in the installation location, use the environment variable CASSANDRA_LOG_DIR to specify the location of the logs directory:

    cd dse-6.8.x
CASSANDRA_LOG_DIR=`<pwd>`/logs bin/dse cassandra

  3. Optional: If you plan to use DSE Analytics, you can use either the default Spark data and log directory locations or define your own custom directory locations:

    Default directory locations

    When you enable DSE Analytics, the Spark components write temporary data and logs to specific directories unless otherwise configured in spark-env.sh or dse.yaml. Make sure these directories exist and have the correct ownership before starting DSE.

    Run the following commands to create and change ownership for the default Spark data and log directories:

    sudo mkdir -p \
  /var/lib/dsefs \
  /var/lib/spark \
  /var/lib/spark/rdd \
  /var/lib/spark/worker \
  /var/log/spark \
  /var/log/spark/master \
  /var/log/spark/alwayson_sql
    sudo chown -R cassandra:cassandra /var/lib/dsefs /var/lib/spark /var/log/spark

    If you plan to run DSE as a different user, replace cassandra:cassandra with that user and group.

    Custom directory locations

    To store Spark data and logs in custom locations, create and change ownership for your custom Spark data and log directories, replacing CUSTOM_DIRECTORY with the absolute path to your custom directory locations. For example:

    sudo mkdir -p \
  CUSTOM_DIRECTORY/dsefs \
  CUSTOM_DIRECTORY/spark/rdd \
  CUSTOM_DIRECTORY/spark/worker \
  CUSTOM_DIRECTORY/spark/log/worker \
  CUSTOM_DIRECTORY/spark/log/master \
  CUSTOM_DIRECTORY/spark/log/alwayson_sql
    sudo chown -R cassandra:cassandra \
  CUSTOM_DIRECTORY/dsefs \
  CUSTOM_DIRECTORY/spark

    If you plan to run DSE as a different user, replace cassandra:cassandra with that user and group.

    Then, update the following environment variables in the spark-env.sh file to point to your custom directory locations. For tarball installations, the default location of the spark-env.sh file is <installation_location>/resources/spark/conf/spark-env.sh.

    export SPARK_WORKER_DIR="CUSTOM_DIRECTORY/spark/worker"
export SPARK_EXECUTOR_DIRS="CUSTOM_DIRECTORY/spark/rdd"
export SPARK_WORKER_LOG_DIR="CUSTOM_DIRECTORY/spark/log/worker"
export SPARK_MASTER_LOG_DIR="CUSTOM_DIRECTORY/spark/log/master"
export ALWAYSON_SQL_LOG_DIR="CUSTOM_DIRECTORY/spark/log/alwayson_sql"

    Finally, in the dse.yaml file, update the location of the DataStax Enterprise file system (DSEFS) work directory to point to your custom location. For tarball installations, the default location of the dse.yaml file is <installation_location>/resources/dse/conf/dse.yaml.

    dsefs_options:
  work_dir: CUSTOM_DIRECTORY/dsefs

Start DSE

  1. From the directory where you extracted the tarball files, run the following command to start DSE as a standalone process:

    DSE_DIRECTORY/bin/dse cassandra

    For other startup options, see Start DataStax Enterprise as a standalone process.

  2. Verify that DSE is running:

    DSE_DIRECTORY/bin/nodetool status
    Result
    Datacenter: Cassandra
=====================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
--  Address    Load       Tokens  Owns    Host ID                               Rack
UN  127.0.0.1  82.43 KB   128     ?       40725dc8-7843-43ae-9c98-7c532b1f517e  rack1

Next steps

Was this helpful?

Give Feedback

How can we improve the documentation?

© Copyright IBM Corporation 2026 | Privacy policy | Terms of use |  Manage Privacy Choices

Apache, Apache Cassandra, Cassandra, Apache Tomcat, Tomcat, Apache Lucene, Apache Solr, Apache Hadoop, Hadoop, Apache Pulsar, Pulsar, Apache Spark, Spark, Apache TinkerPop, TinkerPop, Apache Kafka and Kafka are either registered trademarks or trademarks of the Apache Software Foundation or its subsidiaries in Canada, the United States and/or other countries. Kubernetes is the registered trademark of the Linux Foundation.

General Inquiries: Contact IBM