In this exclusive featured interview with Mohit Lad, CEO and co-founder of ThousandEyes, we learn about how performance monitoring and data can impact decision-making for enterprise networks. Lad answers our questions about the company’s SaaS services, new APIs, and how the ThousandEyes solution is different from the rest.
SDxCentral: ThousandEyes seems to have been making some interesting waves on the performance monitoring front. For the readers that may not yet be familiar with you, can you provide a quick overview of what you do?
Lad: With the adoption of cloud, enterprise IT teams can deploy solutions faster, but this substantially increases dependence on networks outside of their control. Moreover, as they re-architect their infrastructure to exit cloud traffic directly to the Internet from branch offices, the Internet essentially becomes their new corporate backbone. For these organizations, traditional monitoring solutions only show a small part of the overall picture required to manage their infrastructure for performance and reliability going forward.
ThousandEyes is built for these enterprises — without clear network boundaries — and provides network performance management and collaboration capabilities that dramatically reduces the time to identify and remediate problems impacting user experience and business processes. The adoption of ThousandEyes by companies such as Equinix, Flextronics, ServiceNow and Twitter, as well as some of the largest global financial institutions and other Fortune 500 companies, highlights the need for a new approach that transcends traditional enterprise boundaries.
And how is this different from other methods of performance monitoring?
Lad: Traditional network monitoring solutions that live within the four walls of the data center fall short in capturing the highly distributed nature and complexity of the modern enterprise. Network paths are now much longer and complex, and include segments out of control of the enterprise. Isolating network faults requires visibility beyond these four walls and can only be achieved by different types of instrumentation, including client devices, the corporate data center, and the Internet.
Modern network monitoring solutions need to be application-aware and able to single out underlying problems that actually impact user experience. This is a major difference over monitoring solutions from the past that missed application context and only provided visibility into the network layer or over individual devices. ThousandEyes collection agents cover endpoints, branch offices, the corporate data center and cloud, using a mix of active and passive measurement techniques.
A great example of how enhanced network visibility can drive better business decisions comes from one of our large financial services customers. This global bank had chronic connectivity issues with one of their ISPs, impacting customers and brokers accessing their web properties. In the past, the bank would have to sit by and accept the repair timeframe provided by their ISP, typically waiting six to seven hours. By using ThousandEyes, they now have a detailed view into their external networks, previously a ‘black box’, and can see these connectivity issues the minute they occur. The operations team now makes educated decisions as to whether and how to reroute traffic around service provider problems. In this case, network visibility leads to a better application experience for online banking customers.
What are the most common use cases for ThousandEyes today? What has been the response to your cloud-based APIs?
Lad: ThousandEyes is commonly used by enterprise IT, network operations, and network security teams to address a range of network issues. For enterprise IT, ThousandEyes enables rapid troubleshooting of network issues across corporate branch offices and transit providers. Network operations teams can correlate application performance with network issues in their data centers and with SaaS providers. Network security teams can also monitor DDoS, DNS, and BGP attacks in real time and collaborate with vendors to minimize service disruption.
Our customers use our APIs to integrate a variety of monitoring data sources. This integration enables advanced dashboards in their NOCs, detailed reporting and correlation analysis across monitoring systems.
With the move to software-defined networking (SDN) and programmable networks, how do you see ThousandEyes adding value?
Lad: SDN is all about making intelligent decisions at the network layer without needing to get involved with implementation level details. ThousandEyes provides accurate, continuous intelligence on the state of the network, so that organizations can drive better decisions. Rather than hard coding logic into network components, the use of detailed performance data allows network operators to build dynamic logic that improves fault tolerance, reduces downtime, improves performance, and saves on costs.
What specific new use cases could you see by combining SDN with the performance data from your systems?
Lad: I think the most interesting use case is making dynamic routing decisions with real-time performance and event data. In this case, an SDN Controller makes decisions about optimal external network routes, taking into consideration performance, cost and availability data. ThousandEyes provides detailed interface-by-interface performance and path topologies to drive the decision making. Therefore, a way to use ThousandEyes in an SDN model would be to publish performance data through a northbound API that an SDN Controller could use to make routing decisions. The SDN Controller would then communicate to core and edge routers to propagate updated forwarding and routing policies. This sort of intelligent decision making can be used to adjust use of primary or backup links, optimize costly MPLS and VPN links, or reroute around degraded networks.
And why is this hard to do without ThousandEyes? What’s unique about these use cases?
Lad: In order to make optimal routing decisions, you need to know the health of your own network as well as the health of key peering and transit networks all the way to the destination. Visibility into external network links and nodes is crucial in making routing decisions at the edge. ThousandEyes provides this visibility across networks, including hop-by-hop performance data, BGP changes and traffic paths that drive granular routing decisions. And ThousandEyes is packaged with a rich UI for ad-hoc analysis and a robust API to power integrations with systems such as SDN.
What’s your view of the evolution of SDN in the next few years? And what role will you continue to play?
Lad: It’s still early days for SDN as a new approach to manage and configure networks. Adoption will continue to be driven by proprietary solutions from major vendors like Cisco and VMware, startups such as Big Switch and Plexxi and the OpenDaylight open source community. I expect to see a slow pace of consolidation of these vendors happening this year, similar to Cisco’s Tail-F acquisition in 2014. In the next two years we will see adoption continue to be focused on carriers and web-scale network operators.
ThousandEyes will continue our focus on improving the network diagnostics and performance data from enterprise and wide area networks. With higher quality data and more intelligent analytics I believe that the stage will be set for a very compelling SDN value proposition over the coming years.