How SD-WAN can build an application-aware network
Network infrastructure has always been good at moving around bytes of data, but it hasn’t been so good at understanding what that data constitutes. That has made it hard to adapt the network to changing usage, which – when combined with limited bandwidth – can make it hard to match network performance with application requirements.
Past efforts have circumvented the disconnect between application and network by using packet or virtual LAN (VLAN) tagging techniques backed by the use of arbitrary quality of service (QoS) tags.
However, these tags were invariably set by the manufacturer and force users to juggle a completely nomenclature that included new mappings for every application in use. They were also generally overbroad, providing a class for voice over IP – for example – but not allowing the prioritisation of one mission-critical application over another.
In today’s online and cloud-based world, the lack of specificity around QoS controls has meant that employees using the Facebook app on their smartphones probably get the same network priority as project teams using videoconferencing to collaborate with peers in other offices.
Software-defined WANs (SD-WANs) offer a significant improvement to the status quo by giving network administrators a much higher degree of granularity when it comes to managing application performance.
This allows the establishment and enforcement of meaningful, application-aware service level agreements (SLAs) based on particular applications – which can be monitored through the SD-WAN architecture to ensure the SLAs are met.
Because it operates at the software layer and interacts with software-defined network (SDN) architectures, SD-WAN technology can play a far more proactive role in monitoring the performance of specific applications. This allows network administrators to, for example, slow delivery of recreation videos to a trickle during work hours while prioritising Zoom, Skype for Business, or GoToMeeting video collaboration traffic.
SD-WAN based architecture also improve the network’s resilience in the face of unexpected issues, such as when application performance suffers from external factors such as local outages or peak-hour surges in demand for local bandwidth.
In such cases, an SD-WAN architecture can seamlessly communicate with a SDN environment to commission additional bandwidth, or to reroute traffic around congested network segments.
Because SD-WAN controls propagate across multiple infrastructure providers, bandwidth and QoS preferences extend from one side of the network to the other – even through third-party networks and gateways.
In such an environment, the problem that network managers face is no longer ‘I don’t have enough bandwidth’; it becomes ‘the network we’ve got is not the right one’.
The solution for the latter problem is different than that for the former – but broad adoption of SD-WAN simplifies the process of developing and enforcing meaningful application-based QoS practices that keep the business running smoothly no matter what happens.
Request for a demo to see how SD-WAN’s QoS can prioritise different traffic over your network here: www.nttict.com/sd-wan