A new cloud SaaS company has just dropped out of stealth mode, Boundary, a provider of real-time network monitoring-as-a-service.  The company just announced a $4 million Series A round of funding, and they are ready to talk.  They have put together quite a team, and their technology is headed into the cloud.  To find out more about this new company, its technology, and where they are headed, I sat down and spoke with its VP of Marketing, Bob Quillin.  Here is our conversation:
 

VMblog.com:  Now that you have emerged from stealth mode, can you tell us how Boundary got started?

Bob Quillin The emergence and growing popularity of cloud-based infrastructure makes operational challenges acute, especially for enterprises unfamiliar with the high rate of change that comes with the dynamism of the cloud, the tremendous increase in volume of data, and the demand for real-time analysis. The infrastructure operations tools in common use, even those seen as cutting edge, are warmed-over versions of 20-year old tools.  

We (Boundary co-founders Benjamin Black and Cliff Moon) are all too familiar with this because we have designed, built, and operated massive infrastructure using those tools. We have applied that experience to the problems in monitoring, operating, and optimizing infrastructure and have discovered and developed superior techniques and tools.

Benjamin Black is a recognized expert in large-scale technology infrastructure design, implementation, and operation. Prior to founding Boundary, he was an engineering director in the ECN group at Microsoft Windows Live, where he helped build one of the largest content delivery networks in the world. As the manager of website engineering at Amazon, he lead the design and implementation of a new, layer 3 network for Amazon's production network. As principal engineer for Amazon infrastructure, he co-authored the first documents on what later became EC2. As senior manager for information security, he owned and successfully delivered Sarbanes-Oxley and PCI DSS compliance for Amazon infrastructure, including the critical payments processing systems. His work in the Internap R&D team resulted in his being awarded 3 patents in Internet routing and traffic optimization. Black is also co-author on several IETF RFCs on MPLS control protocols and IPv6 network operations. 

Cliff Moon is a builder of and frequent speaker on high-performance, scalable web applications. Prior to founding Boundary, he was the lead engineer for front-end systems at Powerset (acquired by Microsoft, Powerset is a foundation for the Bing service) where he was instrumental in the design,  implementation, launch, and operation of many of the company's production services. Moon is an active contributor to open source projects, and is a highly regarded member of the Erlang community.  

VMblog.com:  Great background information, now if you would, explain to readers what it is that Boundary does, what will it be remembered for?

Quillin:  Boundary provides a real-time, SaaS-based solution for high-volume, high-resolution monitoring of cloud environments. Boundary is designed for organizations whose businesses depend on continuous, real-time operations. Boundary delivers: 

• Real-Time Traffic Flow Monitoring
• High-Volume Analysis
• One-Second Granularity
• Seamless DevOps Integrations
• Affordable Pricing

The monitoring-as-a-service (MaaS) solution is appropriate whether you are running your business in the public cloud, a private cloud, or a hybrid cloud, Boundary's SaaS offering enables ops teams to measure and analyze network and application traffic in any cloud infrastructure, including public cloud providers like Amazon EC2. Boundary's capabilities help you quickly diagnose: 

• Network Traffic Spikes
• Server-to-Server Application Flow Issues
• Bandwidth Usage Hotspots

VMblog.com:  Why do consumers need this type of technology now?

Quillin:  Traditional tools face tough challenges trying to manage and monitor a new world where public cloud hosted infrastructure is growing, hybrid cloud (mixed private and public) is becoming the new standard, and real-time operations are the norm.  In particular, many organizations have adopted a DevOps model based on concepts of continuous deployment and Agile processes.  These organizations are continuously deploying software to their cloud infrastructures - often multiple times a day - but they have no continuous monitoring to support that volume and velocity.  Instead they live with expensive, complex monitoring using legacy tools and hardware probes or reach for open source tools (e.g., Nagios, Ganglia, Munin) resulting in intermittent measurements, slow response, and delayed recovery.   Hardware probes that monitor network and application traffic cannot work in the cloud because all you control is the server instance you deploy. 

VMblog.com:  And how does this help organizations that are moving to the cloud -- like Amazon EC2?

Quillin:  When you deploy to Amazon EC2 for example, all you have access to is your virtual server instance (AMI).  You can't turn on NetFlow on Amazon's routers or deploy a hardware probe in that infrastructure.  We've developed a server-instance resident meter that measures per-second network and application flow traffic without any sampling.  It's low impact but high-resolution.  The meters communicate securely using Transport Layer Security (TLS), a cryptographic protocol that provides authentication and encryption across web connections. The TLS handshake establishes a secure session between Boundary and each meter using a digital certificate exchange that ensures that each meter can only connect with an organization-specific Boundary collector. 

The low-latency (second-by-second), high-volume (thousands of flows, millions of metrics processed per second) helps cloud customers pinpoint: 

• Availability Zone-to-Zone connectivity issues
• Region-to-Region flow problems
• Database, memcached, app, and web server intercommunication hotspots 
• Disappearing and degraded instances
• Instance-to-Instance or Server-to-Sever connectivity issues
• Unreachable instances
• IP address instability
• Degraded EC2 instances

VMblog.com:  Ok, I have to ask.  How does Boundary do it?  What's the secret sauce?

Quillin:  Firstly, Boundary's compact and efficient meters measure per-second flow metrics from packets captured off a server interface. Meters work in the cloud (e.g, Amazon EC2) as well as your private data center - across virtual or physical systems. This information is stored in a standard IPFIX format and transmitted to Boundary's collection services with minimal impact to server resources. Systems supported include Ubuntu 10.04, CentOS 5.x, Mac OSX 10.6 Snow Leopard, and others.  Boundary meters collect flow data such as source and destination IP Address, MAC Address, port, and protocol, ingress & egress traffic volume by packets and bytes, and geo location information. 

This per-second data securely streams through Boundary's high-performance event processing system.  The architecture was designed to be multi-tenant and secure from the ground up and leverages horizontal scaling & automated operations to ensure continuous availability and elasticity. Complex event streaming, NoSQL and BigData analytics enable continuous real-time collection, processing, & display. 

This data then stream out to customer dashboards providing real-time analytics with cutting-edge, live visualizations. They allow you to easily drill-down to pinpoint problems fast.  Real-time views and historical forensics are available at second, minute, and hour granularity.  Automated and customizable annotations allow you to indicate important events related to flows. 

VMblog.com:  So what's next on the horizon for Boundary?

Quillin:  Boundary's public Beta plans to launch in December with general availability targeted for Q1 2012.  Going forward, Boundary will continue to broaden out the inbound metrics it measures and deepen its support for community & partner development to support Boundary API adoption.  The current API is quite powerful already.  It allows you to easily extend Boundary's dashboards with API-based access to customize and integrate our real-time data and analytics into your own dashboards and reports using APIs for:

• Streaming
• Annotations
• Snaps
• Filter Generation
• Meter Management

Jenkins, Chef, and Puppet integrations easily plug into your existing DevOps processes. Automatically annotate your dashboards with each Jenkins deployment to correlate deployments with degradations. Integrate meter deployment with your existing Chef cookbooks and Puppet recipes.

##

Once again, I'd like to thank Bob Quillin, VP of Marketing at Boundary, for taking time out of his schedule and speaking with VMblog.com about his company's launch out of stealth mode.