40- and 100-Gigabit Ethernet (GE): Virtualized servers will generate more traffic than 10 gigabits per second (Gbps) connections can handle. Currently, only a few data centre switch vendors support 40Gbps ports. In 2012, we will see most major NEMs offer higher speeds in their top-of-rack and end-of-row switches, and major server manufacturers offer 40GE interfaces in their high-performance servers. Development will be largely fuelled by the availability of low-cost QSFP+ interfaces. Look for 100Gbsp ports on new offerings for links between core switches. Testing traffic and application delivery at 40/100Gbps speeds (over various distances) will be required to ensure new hardware implementations can meet scalability and quality of service (QoS) requirements.
Converged fiber channel and Ethernet networks: The widespread use of low-cost 10Gbps Ethernet networking, coupled with new data centre bridging (DCB) and Fiber Channel over Ethernet (FCoE) technologies, is finally making LAN/SAN convergence and its economic benefits a reality. Thus far, only Brocade, Cisco, HP and Mellanox have announced FCoE support on their data centre switches. In 2012, we expect to see many additional switch vendors jumping on this trend with FCoE offerings of their own. Conformance and performance benchmark testing will play a significant role in differentiating switch equipment based on overall scalability, traffic prioritization, low latency and storage I/O performance.
Internet security: Zombies – compromised computers ready to act as robot attack systems – are ticking time bombs throughout the Internet. For example, the recent attacks on the WikiLeaks servers in Sweden demonstrate network vulnerability. The latest unified threat management (UTM) devices include multiple security mechanisms: firewalls, intrusion detection and prevention systems (IDS/IPS), anti-virus software, anti-spam and URL filters and VPN gateways. Ixia predicts expanding distributed denial of service (DDoS) attacks will cause NEMs, service providers and enterprises to validate that their network security is air-tight, up to date and doesn’t impact network performance.
Cloud computing: In 2012, Ixia predicts that cloud providers will start to compete on more than price, highlighting reliability, security and responsiveness. To do this, they will need to test their overall compute, storage and network scalability. Cloud providers need to ensure their infrastructure is ready to offer high availability for all customers with different QoS levels for different tiers of service. Enterprise customers are also looking for ways to validate that their cloud providers are meeting service level agreements (SLAs) and shortening schedules to deploy applications in the cloud.
Only by testing all aspects of cloud data-centers can one expose bottlenecks and optimize cloud infrastructures. Uncovering sources of performance degradation include: 1) testing data centre storage I/O in converged LAN/SAN environments, 2) access and aggregation switching tiers at 10/40GE, 3) core switching tiers at 100GE (data-centre interconnect), 4) virtualization layers with application workloads, and 5) security effectiveness and accuracy. The consumers of cloud services will also need test solutions that can advise how much and where compute resources are needed to deliver adequate user quality of experience (QoE).
IPv4/IPv6 co-existence: On 1 February 2011, the Internet Assigned Numbers Authority (IANA) allocated the last freely-available block of IPv4 addresses, pushing IPv6 addresses to the forefront. Service providers and enterprise are under the gun to prepare their networks for the influx of IPv6 traffic. Two major catalysts for IPv6 implementation are the large number of government-driven policy initiatives (US, Australia, Czech Republic, Germany and many others) mandating IPv6-ready hardware and software deployment across public agencies, and Organization for Economic Co-operation and Development (OECD) countries such as the Netherlands providing financial incentives for organizations to accelerate the IPv6 deployment. Service provider investment in upgrading access networks to support both IPv4 and IPv6 will continue through 2012. In the latter half of 2012 we also expect to see many large enterprises rolling out IPv6 on corporate networks. To ensure this evolution is transparent to networks users, NEMs, service providers and enterprise IT will engage in both public and private tests that demonstrate equipment readiness.
Long Term Evolution (LTE): By the end of 2011 there will be more than 30 commercial LTE deployments and many more trials ongoing world-wide. 2012 will see a doubling of commercial LTE networks, with North America and Asia leading the way in early deployments. As smart phone adoption rates increase and the use of media-rich applications expands, mobile operators will see continued capacity challenges. 2012 will see operators use multi-pronged approaches to solving capacity issues including: 1) deployment of more spectrum-efficient technology, 2) increased data offload using Wi-Fi and small cells, 3) improved network traffic management, and 4) new business models and charging plans. With these approaches, operators hope to not only increase raw capacity, but to lower the delivery cost per bit. Operators are spending billions to make their network more efficient, and prudent operators will thoroughly validate their networks in order to deliver on these objectives. The realistic simulation of millions of clients using media-rich applications will be a key driver for performance and capacity assurance. The very first implementations of voice on LTE networks using Volte will begin to occur in 2012, but don’t expect wide deployment until 2013 or beyond. 2012 will see mobile data networks become a more fertile ground for security attacks, causing wireless security testing needs to also ramp up.
Evolved Packet Core (EPC): The EPC is the aggregation point for all traffic originating from multiple wireless access types, including LTE and different variants of 3G technology. In 2012, “scalability” will be the mantra for new Epics. These core networks must handle massive amounts of converged voice, video and data traffic on a single IP-based network. The network and its component devices must prove their scaling capabilities in multiple dimensions. In particular, control plane scalability will be a major concern for NEMs and operators in 2012. Poor performance and faulty deployment plans can have severe consequences. History serves as evidence that subscriber dissatisfaction from a sluggish network or service outages is a real threat. The complexities of performance and how to properly manage network traffic are far too great for guess work. In 2012 and beyond, successful NEMs and operators must give due diligence to these critical issues and seek validation from test companies that are experts in this space.
Mobile backhaul: Driven by bandwidth-hogging multimedia and mobile data, mobile operators are actively transforming their legacy TDM backhaul networks to a cost-effective IP-over-Ethernet paradigm. Clock synchronization is no longer a “barrier” to IP/Ethernet backhaul. The functionality and inter-vendor interoperability of IEEE 1588v2 precision time protocol has been proven in multiple public industry tests over the past 2 years, and 58% of service providers plan to deploy it by 2013 (source: Infonetics). The last hurdle for service providers is benchmarking the performance of boundary and transparent clock implementations, at scale, in the context of real-world traffic delivery. The onus will be on NEMs to prove the performance of their backhaul switches with tangible, repeatable test results. Service providers will also engage in pre-deployment testing to determine how to optimize network configurations to accommodate network growth and SLAs.
MPLS-TP: In 2012, the industry will turn to MPLS-TP performance testing to validate critical measurements such as scalability and automatic protection switchover. Public industry tests will also focus on MPLS-TP features that equip service providers with more advanced network management. Ixia expects to see very few service providers deploying MPLS-TP on their networks in 2012. In our opinion, widespread deployment is still at least 18 months away because only a handful of vendors have participated in public interoperability events – indicating an immaturity of many vendor implementations. Furthermore, proving MPLS-TP as a viable technology is only the first milestone toward widespread deployment. Service providers want evidence that MPLS-TP delivers service at scale with guaranteed QoS and “five 9’s” reliability.
Rich Media: A perfect storm is developing for network operators. Smartphone users expect information and entertainment 24 hours a day, in full living color. Wireless data usage will continue along its anticipated exponential growth curve, and large amounts of inexpensive bandwidth are no longer available. Enterprises will continue to move to the cloud, and critical applications will depend on guaranteed network access and performance.
We predict that network operators will need to compete and live up to their reliability commitments. They will need to maintain customers by ensuring that all types of customer service levels meet expectations and contractual requirements. Pre-testing network upgrades and advances prior to initial deployment is crucial to ensure readiness for anticipated customers’ usage. This type of testing will involve real-world subscriber modeling that simulates a large quantity of mobile, home and enterprise application usage – web, email, streaming video, VoIP – in typical and unusual scenarios.