Optimising The Cloud with Network Intelligence
By Sean Ong, Country Manager, Brocade Malaysia
The rise of multimedia and the growing use of real-time applications—including voice over Internet Protocol (VoIP), video, desktop sharing and live webcasts—have placed heavy demands on network bandwidth and performance. The capability of the network to respond to changes in traffic patterns or congestion is automatically becoming more important for organisations to stay competitive.
That’s because application delivery and workload hosting rely heavily on the network. Which is why, In the Brocade Global CIO Survey 2015, 60% of CIOs mentioned network performance as a key challenge for cloud services.
Achieving Visibility, control and automation through Network Intelligence
The world of IT has changed dramatically in the last 20 years, but the underlying networks and the way data centres are constructed have not changed with it. IT professionals are still grappling with basic questions like: “Do I have enough bandwidth/capacity in my network?” or “Can my network automatically redirect traffic when there is congestion?” While there have been some solutions to address those challenges, none has been satisfactory.
CIOs today need to have greater understanding on what happens on a network. But network intelligence is more than just visibility and control. It has to include automation. This was promised in the past but most offerings were vendor-centric, complex, limited in functionality and expensive to deploy.
With the onset of Software-Defined Networking (SDN), it is now possible to realize true automation. SDN applications that work with an OpenDaylight-compliant controller can enable intelligent flow management and increased network agility on an open SDN architecture, for real-time network visibility and policy enforcement.
New services can be introduced with increased speed and value-added Service Level Agreements (SLAs), while open SDN architecture accommodates low risk, open and modular solutions.
SDN-based policy detection and management enables informed decisions to be made, based on real-time proactive network visibility. It also allows network managers to mitigate network flood attacks, throttle bandwidth-intensive workloads, reprioritize rogue traffic, and redirect suspicious activity for further analysis.
Automated Flow Management
Automated flow management, should be scalable and capable of supporting multiple aggregated 200 GB flows simultaneously, enabling data centers to handle large volumes of traffic flows. It should also be user-friendly, with an interactive web-based GUI for event logging and traffic statistics. And it should offer a range of flexible options to respond to traffic changes, and be able to remark, redirect, rate-limit, or even drop traffic flows. This ultimately adds more value and enables additional new SLAs, while supporting new services faster with bandwidth scheduling and flow-based traffic mirroring services.
Network Visibility and Traffic Analysis
How does the SDN application know when something is happening in the network? This can be accomplished through a feedback mechanism like SFlow.
SFlow provides enhanced network visibility & traffic analysis; contributes relevant data to an overall network security solution; and can be used for SLA verification, accounting and billing purposes. SFlow has been implemented in network switches & routers for many years and is now often implemented in end hosts.
The network can respond by re-routing large flows, dropping flows that match Denial of Service Attacks (DOS), duplicating the flow to an analytics platform for further analysis, or simply by throttling down traffic to specific websites. All these can be done automatically from a central point by the SDN application.
Network Automation – Start Putting Your Network to Work
Network automation has always been a challenge for IT administrators due to the difficulties of having network visibility, control and automation of different network devices; especially in a multi-vendor environment. Very often, the existing solutions offered are more reactive rather than proactive, and by the time the IT administrator is alerted to issues, it would already be too late.
The answer to this problem is to have a more open and programmable network that enables not just automation but innovation as well. A variety of SDN applications can be developed, since the interface to the SDN controller is standard.
This is an attribute of the New IP which is designed to meet the needs of cloud, mobile, social and big data, where the demands of IT infrastructure agility and automation is key to helping organizations remain competitive and stay relevant.