Jump to: navigation, search


Blueprint and bugs info: https://launchpad.net/savanna/0.3/0.3

Elastic Data Processing (first release):

  • MapReduce jobs could be created and launched using REST API or UI including Pig and Hive scripts in flexible, configurable manner;
  • jobs could be stored in both Savanna internal DB and Swift;
  • Swift supported as Data Source for MapReduce job input and output;

Hadoop provisioning:

  • data locality is fully supported now - rack and 4-level node group awareness for both HDFS and Swift;
  • Vanilla plugin supports parallel provisioning of latest Hadoop 1.2.1 clusters including Oozie and Hive that are needed for EDP;
  • Hortonworks Data Platform (HDP) plugin supports HDP 1.3.2 provisioning in parallel manner too;

Core Savanna:

  • Extended network configuration including floating IPs pool configuration for node groups and auto assignment support;
  • Neutron support (nova-network is supported too);
  • all DB-related code is now moved to the “conductor”;

Python client:

  • all core and EDP features are implemented and fully operational through the python-savannaclient;
  • it supports both v2.0 and v3 APIs of Keystone REST API;

OpenStack Dashboard plugin:

  • all core and EDP features are supported in UI;
  • it uses python-savannaclient for interop now;


  • supports building images with Oozie, Hive and optional MySQL for Vanilla plugin;
  • one-click image building for Vanilla plugin;
  • UI integration tests implemented using Selenium;
  • Savanna could be installed in Fedora using yum;
  • all four main Savanna subprojects are available now at pypi.