ThousandEyes Overview

Editor's Notes

• #2: Three questions: Who is my audience? Where is their mind at? What are they thinking? What’s my objective from the presentation?
• #7: What we’re moving from is this controlled environment where you own everything, to an environment where there are many different segments: Enterprise segment, internet segment, hosting segment and end users. Challenges having this sort of view are multi-fold: Because you can’t see the entire environment, it’s very hard to troubleshoot problems that have happened. A lot of time you’re looking at things in silos – so you’re either looking at the network on its own, the application on its own. So any time there is an app issue – lots of finger pointing going on. The other interesting piece in this world, is the exercise of troubleshooting – you’re no longer working just within your org – working with Salesforce, Office365; finger pointing or troubleshooting across organizations becomes very difficult. What we found is a big gap in the industry – the ability to visualize the big picture AND a strong collaboration capability – to help left side of picture collaborate with the right side. At ThousandEyes we approach this differently. We treat these different network segments as ONE network perimeter that you can troubleshoot across, whether an application consumer or provider. Additional background - existing technology limitations: Limited to silos, not solving end-to-end visibility Focused on reporting on symptoms, not root cause Built for controlled environments, not dynamic, distributed and changing environments Dependent on complex application specific instrumentation Deployments can take months and years
• #8: The question we asked ourselves – the big idea - was, what if we can build a view that does not differentiate these network segments, which can actually give all parties the view of exactly what's going on in the entire network? That treats this all as one coherent network? That lets you: First troubleshoot problems really fast. Monitor for issues before they impact applications, customers, services. Plan and make better decisions when you’re trying to change your architecture, build cloud applications and so on. That’s the premise behind ThousandEyes.
• #9: As well as visibility, the ThousandEyes platform provides situational context and awareness to give you more actionable insights. Three core tenets of the platform are: Deep Path Analysis: view networks outside of your control, with a similar level of visibility you have in your own network. X-Layer: correlates application performance to underlying infrastructure metrics and views. Why is an application misbehaving? Is it a delivery issue or which component of the network or infrastructure is responsible. Interacting Sharing: in this new world, you need to collaborate beyond organizational borders to solve problems – to collaborate on an issue with people (internal, external) who are responsible for solving it. Let’s jump into a demo and start looking into the product and look at what sorts of information we’ll collect.
• #10: How do we give you a consolidated view on what’s going on in the entire environment? Instead of trying to figure out how to instrument apps, the philosophy is to deploy smart agents that can figure out what the dependences are to any application – whether inside or outside your environment. The fundamental difference between ThousandEyes and some of the other solutions is that we’re application agnostic, we don’t really care where the application sits. Since we are a cloud service, we consolidate the data sources to provide you with a really nice view of what's going on in the Internet and in your own environment. Take for example an Enterprise IT environment with branch office locations. You may be getting complaints about slow performance from certain regions, but you don’t really know when, where and why your ERP application is slow. With ThousandEyes, you can deploy Enterprise Agents – very easy to deploy software agent that you can push in each of your branch offices. With this you get a view on performance inside your environment to any applications. Enterprise Agents can be installed in multiple locations to get a mesh of performance measurements that supports traffic like VoIP and Unified Communications. This provides visibility into the MPLS or VPN links between enterprise sites. TRANSITION Enterprise Agents can also test to targets outside of the organization, such as hosting or SaaS providers. So if Salesforce or Office 365 performance is important to measure, you can instrument your key branch offices and call centers. TRANSITION For companies that deliver an online application, we also provide valuable visibility. It could be a cloud based application that’s B2B or it could be an e-commerce application. ThousandEyes provides the ability to look at what’s going on in the entire internet and how it’s affecting the application. For this use case we use Cloud Agents – which are deployed across different parts of the internet and are managed by us. They collect intelligence from dozens of metro areas and networks and how an application is being delivered to those regions.
• #12: Verisign is a major internet infrastructure provider, with Managed DNS, registry for .com and .net top-level domains, and DDoS mitigation services. Verisign uses ThousandEyes to solve challenging infrastructure issues at large scale, including monitoring of Anycast DNS and performance of two of the DNS root servers, which they were not able to easily isolate before using ThousandEyes.
• #13: Headquartered in San Francisco, Zendesk provides an integrated on-demand help desk and customer support portal solution based on the latest Web 2.0 technologies and design philosophies. Zendesk delivers their solution in the form of Software as a Service (SaaS). As of 2013, more than 30,000 companies rely on the Zendesk platform to provide customer support. Zendesk would encounter situations where our application was working well but we would still hear customers report slow performance.. In May 2013 ThousandEyes alerted availability problems to one of the Zendesk services. This coincided with Zendesk customers reporting performance issues confirming that this event was affecting some of Zendesk’s customers. The operations team at Zendesk could not find anything misbehaving on their application stack, raising suspicion on a hosting provider Zendesk used. In the past, when such problems had occurred, the operations team at Zendesk was not able to identify issues outside the application stack. With ThousandEyes, the operations team could use Deep Path to identify specific routers that were dropping packets, all belonging to the same Internet Service Provider (ISP). Zendesk could see that the ISP in question connected directly with Zendesk’s hosting. This was a network issue at the edge of their hosting provider which was affecting Zendesk’s application performance. With the Interactive Sharing capability, Zendesk shared the view identifying the problem router with their provider and the provider was able to immediately react to the problem.
• #14: Since ServiceNow is a SaaS application handling critical IT functions within the enterprise, it is crucial for them to ensure their customers do not run into availability and performance issues. In Feb 2013 ThousandEyes alerted ServiceNow to degradation of their web performance to one of their data centers. By looking at the network layer of their test, the operations team instantly identified it as a network issue. Using the Deep Path Analysis (DPA) technology, they further isolated the issue to a specific upstream provider. The Director of Network Engineering immediately knew where the issue was and reacted by first using the interactive sharing capability in ThousandEyes to share the problem with the affected provider, then immediately changed the BGP (Border Gateway Protocol) preference to prefer another provider at the same location and route around the problem. He was able to actually visualize the BGP change and the change in the flow of traffic through ThousandEyes to ensure our change worked and we could resolve this issue in no time. ThousandEyes’s X-Layer technology gave ServiceNow a single solution to examine the correlation between the HTTP layer, IP layer as well as BGP layer.
• #15: Twitter identified large performance improvements in the lead up to the 2014 World Cup. With so many expected users in Brazil, Twitter was testing performance and found poor DNS resolution times in South America. Using ThousandEyes they quantified the performance issue, identified a faster location to host DNS for Brazilian users and cut resolution time from 300ms to 60ms, leading to a dramatically better user experience for mobile users.
• #16: ThousandEyes has capabilities to help you with everything from in-the-trenches troubleshooting to ongoing outage detection, trending analysis and network optimization.
• #17: ThousandEyes is best known for troubleshooting. Deep Path Analysis, shown in the Path Visualization here, makes it easy to understand why network performance is degraded and where the problem is located. Interactive Sharing then helps you resolve internal or service provider incidents faster by making a point-in-time snapshot or live feed of data available for collaborators in your organization or in your vendors’.
• #18: Before you even begin troubleshooting, you’ll want to be alerted when a problem is brewing. ThousandEyes has highly customizable alerts that aim to reduce false positives so that you’ll only be notified when an event really needs your attention. And our integrations with PagerDuty, via Webhooks or via our well-documented API make it simple for you to use your existing investments alongside ThousandEyes.
• #19: Whether you are trying to report on improved performance, an outage or just dig for optimization opportunities, reporting is a powerful way to communicate performance trends. ThousandEyes Reports are built with customizable widgets, have multiple types of visualizations and can be scheduled, saved or shared to create a record of your network.
• #20: Use detailed Path Visualization and hop-by-hop metrics to optimize VoIP performance, try out a routing change on a small portion of your network, plan for a new data center or test a roll out of a new application or SaaS service. With synthetic testing you can plan and optimize before you fully implement changes on your network.
• #21: Our customer success team is your team that takes a collaborative approach to keep your network at its peak performance. Staffed by network experts with deep knowledge of enterprise and internet environments. We empower you to get your job done more efficiently and effectively. From routing issues to DNS misconfigurations and DDoS attacks, we work with your team to troubleshoot tough problems and find the root cause. That’s why we’re here. Looking for best practice tips? We’re happy to share. Have a product suggestion? Bring it on. We’re committed to getting you the best network performance.
• #22: Getting started is easy. Sign up in under a minute; subscribe to any one of 18 of the top SaaS apps you’re interested in; deploy enterprise agents if you’re interested; leverage our customer success team. To make ThousandEyes even more useful to you, we can configure it to your needs. From what you’ve heard, what do you want – what’s interesting? Will you use enterprise agents, will you use cloud agents, etc. To set up a sales trial. We also have full docs, videos, chat – we’re here to help you get started!
• #23: Collect, integrate, and analyze data from all the network segments