Open Flow : Based Server Load Balancing Gone Wild

* Opengear supports the OpenFlow/SDN Interoperability Lab. This Software Defined Networking (SDN) technology from the Open

Brocade is fully invested in making SDN completely modular to deliver New IP networks based on an open ecosystem. The Brocade SDN Controller is continuously built from the OpenDaylight code, with a virtuous cycle of contributions back to the OpenDaylight community. Physical and virtual networks from multiple vendors can be managed within the Brocade SDN Controller, and organizations can gradually introduce ever-larger portions of their existing networks into the controller domain with single-source technical support for the entire

One of the largest goals of the Acme Relocation Project is the design and implementation of the new network infrastructure. Acme Produce 's Management team has requested the Enterprise Network Engineer design and implement a novel network infrastructure. The current onsite Information Technology team of Acme Produce is lacking Senior Network Engineers. Acme Produce 's Management team has decided to outsource this portion of the project to avoid missing project deadlines, as the Senior Network Engineer positions available have been vacant for several weeks.

Software defined networking is a new concept that redefined the current network device architecture and the way they behave. Due to the exponential growth of the Internet and services attached to this worldwide infrastructure, a number of issues have cropped up in modern day networks. This paper discusses a few of these limitations and how the software defined networking approach tries to alleviate them. OpenFlow, released in the year 2008, is a commercially viable implementation of this concept that is being well received by the networking industry at large. The paper also discusses the OpenFlow protocol, its implementation and the challenges facing its widespread acceptance

The SDN Technology states the reference points and interfaces to the controller. A number of functions, which are internal to the SDN Controller, are specified in the architecture. The SDN controller is restricted to features that allow interoperable applications that are to be deployed. The architecture is ambiguous to the protocols across the interfaces.

This section contains information about the implementation of the SDN switch core, interconnection of multiple SDN switches with the SDN controller (PowerPC) and the attacker nodes (Microblaze). The main functionality of the SDN switch is to modify packet header fields based on the flow table and forward it to the next port(s). The SDN controller is responsible for programming the flow table in each switch and monitor these switches to observe each packet flow. The Microblaze processor, acting as attacker nodes plays the role of an outside network and transmits packets at different programmable rates to the SDN switch network using an array of packet drivers. The big picture showing the connection between different components is given in Figure 3.1.

The OFFN-RM project design requires an event listener to accept REST commands, updates controller state, and then construct appropriate action with the OpenFlow protocol. One testbed scenario involves a login scraper that will logically sit on top of the controller, to act as a clearinghouse for security events and to ingest information from devices, security mechanisms, appliances, identity services, and Globus; the scenario alternative is a direct connection. Customization of the login scraper seeks (a) to identify patterns and thresholds to block security threats, and (b) to detect and steer huge data flows around performance limiting mechanisms, by monitoring Globus GridFTP event logs and mining data transfer flows for multiple socket connections, restart markers, and data channel information. Login scraper customization will involve three events: (1) for events with a high priority, the action will be blocked and event listener code modified for REST commands to block source addresses; (2) for Globus flow events, steering will be accomplished with registration of the five-tuple; (3) for regularly scheduled events, automatic clearing of blocked source addresses will occur after a set time period. Extant hardware in the form of components of a decommissioned cluster (12 PowerEdge R620 compute nodes: dual Sandy Bridge E5-2650 8-core 2.0 GHz, 32 GB RAM) will be deployed for login scraper storage and containment

The DNS is a vital Internet component providing mechanisms for resolving host names into IP addresses and vice versa. In addition, DNS supports other Internet directories concerning DNS name servers and mail exchangers among others (Garfinkel & Spafford, 1997). Despite being a universally accepted and widely used system, insecure protocols and lack of authentication among other factors within the DNS threaten proper functionality of the DNS. Moreover, the accuracy of the information contained within the DNS is vital to many aspects of IP based communications.

Scalability is a major requirement for virtual networks. For this reason, virtual link aggregation will constitute an important requirement for network virtualization. Virtual link aggregation can be defined as a set of virtual links that follow a common path and are similarly treated between a pair of virtual nodes. It can be performed by carrying at least two types of identifiers in the data plane, one for virtual network identification and another for hop-by-hop forwarding. Hence, virtual link aggregation enhances scalability and also improves performance.

The client and server architecture is a distributed application that divide the tasks between the entities that provides the resource or service, called servers, and the entities that request for a service, called clients. Fundamentally, clients and servers are two software entities where one software is fulfilling the request of another software entity. The client machine request for the service and the server machine interprets the request message received from the client machine. In order to fulfill the request of client machine, the server may have to refer various knowledge domains, process data and sometimes make an additional request to another server. The client and server may reside on the same machine or sometimes lie

Software-Define Network (SDN) has emerged as a very promising network architecture in recent years. It largely simplifies the network logic and makes the network traffic more controllable by decoupling data plane and control plane out of the traditional network architecture, and having centralized controllers to control network switches. The significant difference from the traditional network architecture is that the network traffic routing is centralized into controllers such that network switches are only need to perform forwarding, instead of complicated routing protocols. The centralized controlling scheme of SDN has many novel and important capabilities such as global view of the network, software-based traffic analysis, and dynamic reconfiguring of network forwarding rules. The SDN has attracted the interests of many attackers due to the use of centralized controlling architecture and many security flaws are remained to be addressed yet. Distributed Denial of Service (DDoS) attack is one of the most frequent network attack that no effective countermeasure is acknowledged in traditional network. Fortunately, the characteristics of SDN bring us new chances to effectively against DDoS attacks. In this paper, we are going to discuss the classifications and characteristics of DDoS, analyze the advantaged capabilities of SDN architecture, and present a survey of the methods to detect and prevent DDoS attacks using SDN, and review the studies of SDN as a victim of

Abstract - Infrastructure as a service refers to the capacity gave to the purchaser as to procurement preparing, capacity, systems, and other major figuring assets where the buyer can convey and run discretionary programming. In today’s cloud however, the control provided to the end user is very limited. It does not extend full administrative capability like controlling network flows and configuring the network. Therefore, it becomes imperative that there has to be an architecture that provides a certain amount of control to the end-user for modification of network flows. In this project, it is our effort to bring out the various aspects of Software defined networking for providing infrastructure as a service compared to today’s cloud

Software-defined networking (SDN) promises something akin to enterprise networking utopia by separating the control panel from the physical network topology to create an environment where all switches and routers take their traffic forwarding cues from a centralized management controller. In a software-defined networking environment, this software-based controller must have an end-to-end view of network resources and capacity.

Software Defined Networking (SOFTWARE DEFINED NETWORKS) can enormously simplify network management by offering software engineers network-wide visibility and direct control over the fundamental changes from a legitimately unified controller. In any case, existing controller stages offer a "northbound" API that powers software engineers to reason physically, in unstructured and ad-hoc routes, about low-level conditions between various parts of their code. An application that performs different tasks, for example routing, monitoring, access control, and server load adjusting must guarantee that packet processing rules introduced to perform one undertaking do not supersede the usefulness of another. This outcomes in solid applications where the rationale for various undertakings is inflexibly interwoven, making the product difficult to compose, test, investigate, and reuse. Modularity is the way to overseeing complexity in any programming framework and SOFTWARE DEFINED NETWORKSs are no exemption. In the past research, it has handled an imperative extraordinary case, where every application controls its own cut—a disjoint bit of traffic over which the occupant or application module has complete visibility and control. Notwithstanding traffic disconnection such a stage may likewise support subdivision of network resources, for example link data transfer capacity, rule-table space and controller CPU and memory to keep one module from influencing the execution of another, but

Changing traffic patterns: Applications that commonly access geographically distributed databases and servers through public and private clouds require extremely flexible traffic management and access to bandwidth on demand.