Open Daylight Helium Features Fill the Deployment Vacuum

The OpenDaylight community has launched its much awaited software release, the Helium, an upgrade to its former OpenDaylight controller called Hydrogen. A detailed look at the features introduced in Helium makes it clear that the community is targeting to fill the gaps left by the previous deployment in various networks. Many OEMs and Service Providers are embracing the OpenDaylight Hydrogen release to primarily develop Proof of Concept (PoC) solutions.

Hydrogen, the first release of OpenDaylight controller, provided several basic functions on top of a modular and multi-protocol aware SDN controller. Though it was claimed to be a “deployment ready” SDN controller, several key features required for deployments such as policy management, high availability, and a native integration with OpenStack, were missing. This new release from the OpenDaylight community has addressed many of those deployment gaps.

image

Below is a look at the key features introduced in Helium:

  1. Deeper integration with OpenStack – Helium supports OpenStack features such as Security Groups, Distributed Virtual Router and Load Balancing as a Service. Today, SDN & NFV solutions ride on top of each other. There is a growing need to manage Virtual Network Functions (VNFs) and orchestrate them through a cloud controller such as OpenStack. Deeper integration with OpenStack will increase the adoption of OpenDaylight controller by cloud service providers. However, the Helium release carries only a technology preview of the OpenStack integration and hopefully, Lithium, the next major release, will have more to offer.
  2. Improved OVSDB integration – Today, Open vSwitch Database (OVSDB) has become the defacto protocol to manage the configuration of OVS switches in data centers. OVSDB provides mechanisms to configure ports, interfaces and policies in an OVS switch. Helium supports creation of virtual networks and OpenStack Neutron features such as LB as a service and firewall as a service.
  3. Clustering of Controllers and High-availability support – Helium provides high-availability support through clustering. The controller supports synchronization of state/event data across the clustered OpenDaylight controllers, as well as clustering services, which the controller modules can use to get state and event synchronization. The controller also provides a transaction API to maintain transactions across the nodes in a cluster. If a controller instance goes down, the SDN application running on the controller goes down with it. It is the responsibility of the SDN application to ensure its own resiliency by having multiple instances and providing its own state synchronization between the instances.
  4. AAA capabilities - Helium provides the control to the network to decide the permissions and resource usage limits for various users/services. It allows a service provider to integrate the OpenDaylight controller with an external identity services engine or a policy server to provide authentication, authorization and accounting functions.
  5. OpenFlow Table Type Patterns support – Helium provides the ability to define logical forwarding pipelines in the OpenFlow enabled switch. This allows the hardware in the OpenFlow enabled switches to expose more of their capabilities to the OpenDaylight controller.
  6. Policy framework – The new policy management framework introduced in the Helium release allows an administrator to group network devices and apply network policies. This will help in abstracting the network nodes in a large deployment of OpenFlow enabled devices.
  7. PCMM Support – The Helium release proves that OpenDaylight is not just targeted at data centers or cloud service providers but can also be deployed in cable networks. Helium provides a Packet Cable Multimedia (PCMM)/Common Open Policy Service (COPS) southbound plugin. With the support for PCMM modules, OpenDayLight controller can provision CMTS as a network element which manages service flows with dynamic QoS.
  8. Tools for building Service Function Chaining (SFC) - Service function chaining allows administrators to define a series of services through which network traffic must go and in a certain order. For example, you can make the traffic go through a firewall function, a Deep Packet Inspection (DPI) function before forwarding the packet to the next hop in the network. In the Helium release, the community has defined an SFC architecture focused on building Service Chains with Open vSwitch-based Service Functions. It simplifies the process of building service function chains and paths.
  9. And, last but not the least … a much improved Graphical User Interface (GUI), which includes simpler and customizable installation process. The OpenDaylight community has used Apache Karaf container to build the GUI framework. Apache Karaf support allows users of Open Daylight to choose the core features or add-ons they wish to have in their OpenDaylight installation using both command-line or the user interface.

The OpenDaylight community has made significant strides to the OpenDaylight controller with the Helium release, delivering a more usable and useful tool. With the support for features such as deeper integration with OpenStack, improved OVSDB integration, High Availability, and Policy Management, OpenDaylight has become an ideal candidate for deployment in green field SDN architectures.

The major missing element in the Helium release is the support for OpenFlow v1.4 as it is more than a year since OpenFlow v1.4 specification was released and there are not many controllers that support OpenFlow v1.4. Also, Cisco’s OpFlex southbound API support couldn’t be added in-time to the Helium release. Hope, Open Daylight Lithium release will have support for both OpenFlow v1.4 and Cisco’s OpFlex interface.

References:

http://www.opendaylight.org/software

https://wiki.opendaylight.org/view/Main_Page

 

Leave a Reply

Your email address will not be published. Required fields are marked *