condla / hadoop-bootstrap

Role Name

• This role should be used to install and configure your operating system before you plan to install Apache Hadoop and its Ecosystem components.

• It also installs all packages and libraries required to secure the cluster using the Kerberos protocol.

• The first thing it does per default is to install the epel repository, which is needed to install python-pip. Pip is required to install several client tools for Hadoop. If you do not wish to install epel, set the variable install_epel to no in your playbook.

Requirements

• A CentOS-7.x or RHEL 7.x
• all kernel updates already done: this is important since there might be changes in the kernel configuration. If you update the kernel and change its configuration the instance might not boot

Role Variables

Default variables:

• hadoop_db: Defines which db will be used for the Hadoop installation that will be done.

Security Related Variables:

• krb_realm: Define the realm you want to be using for your KDC
• kdc_hostname: configure where your KDC is running
• kdc_admin_hostname: configure where kdc admin instance is running
• krb_domain: define krb_domain

Dependencies

condla/hdp-instances

Example Playbook

Including an example of how to use your role (for instance, with variables passed in as parameters) is always nice for users too:

- name: bootstrap hosts for the hadoop installation
  hosts: hdp
  gather_facts: yes
  become: yes
  roles:
    - hadoop-bootstrap

License

BSD

Author Information

Stefan Kupstaitis-Dunkler ([email protected])

hadoop-bootstrap's Issues

Reboot handler does not reboot machines or throws an error

creating of /etc/hosts file does not scale

Create /etc/hosts file on one node, download it and deploy it on all nodes instead of procedure that is currently implemented