Primary tabs

NetSocket Announces Cloud Experience Manager

4 Jul 2012

Service providers (both network and IT outsourcers) and enterprises are undergoing a major transition to cloud network architectures using a mix of private, public, and hybrid IP clouds to deliver enterprise VoIP, videoconferencing/telepresence, and other UCC services. Hosted communications service delivered to large enterprise typically involves at least three clouds: the enterprise's private cloud, the public cloud and the service provider's private cloud.

From a service management perspective these changes introduce significant new challenges. Historically content, session and IP topology layers in a communications network have been architected and managed as separate entities. This no longer works. Furthermore, quality of service metrics such as packet loss, latency and jitter don't provide reliable information about the end user's actual UCC quality of experience; and certainly do not provide enough diagnostic information to determine the cause and location of quality issues or even to bracket issues to the correct cloud.

Given this set of circumstances, service providers (SPs) need to think beyond network SLAs and make sure service assurance in the cloud realm includes the network, the application layer and the business layer in order to: (1) Provide an overall satisfactory customer experience, and (2) an SLA monitoring capability built to validate external service delivery partner performance by definitively characterizing the role of the enterprise's own components in the root cause analysis. On the enterprise side lack of visibility into the end user's (i.e., employee, supplier and customer) quality of experience leaves organization IT waiting for complaints to come in. IT, however, needs to get out in front of problems by proactively managing end user experience through early detection and rectification of performance issues before they can impact users, be they related to network events or upcoming capacity shortages based on trend analysis.

NetSocket recently announced its Cloud Experience Manager (CEM) solution targeted at addressing these SP and enterprise needs. CEM uses a distributed architecture deployed at major network segments (e.g., datacenters and contact centers) to provide a realtime correlated view of the different layers of the user experience: content quality - pertaining to the voice, video and data stream; session quality - basically the quality of the E2E transaction; and the network quality (a key differentiator) - involving the realtime hop‐by‐hop IP topology related to the physical devices and the endpoints in the network. An integrated view of into end user experience from all three layers is delivered through a single console screen, typically deployed in the NOC, allowing enterprises and SPs to proactively detect, isolate and remediate network issues within UCC environments before they affect end user experience. The management system can be deployed in a VMware virtual machine environment.

The core of NetSocket's technology is its patented IP Correlation Engine (ICE). This multilayer information mashup service provides a dynamic and actionable analysis of service behavior. ICE technology correlates the three primary components of user experience: content, session and IP topology. Additionally, the technology allows organizations to quantify and enhance user experience via realtime monitoring and correlation across these components. These solutions can be deployed both by enterprises and SPs.

CEM provides realtime correlation of content session, and IP topology data of the following types into in a session-specific Quality of Session Record (QSR) for the network operations team:

  • IP network status and hop-by-hop path for each voice, video and data session

  • Correlation of IP network events to sessions

  • Listening and conversational quality MOS (mean opinion score) and R-factor

  • Signaling and media (i.e., RTP, RTCP) statistics

  • Packet drop/loss

  • Jitter

  • Call quality statistics (e.g., success rates, delays, one-way audio, echo)

  • IP flows analysis (i.e., NetFlow, JFlow)

The QSR allows organizations to correlate how disparate elements of their service delivery chain are impacting the quality of end user experience, resulting in an improved ability to prevent performance issues and reduced time needed to isolate and repair performance problems. Additionally, key quality and performance indicators are compared to user-configured thresholds. If these thresholds are exceeded, the solution automatically generates and delivers alerts via interactive dashboards. QSRs can be saved for customizable time periods up to 30 days allowing customers to conduct always-on monitoring while being able to capture and easily isolate realtime and historic performance information at a granular level.

CEM peers with routers in the IP cloud using IGP/EGP protocols (e.g., OSPF, BGP) to learn IP network topology and state. This is a completely passive, "listen only" process that is no different operationally from adding a routing peer to the routers in the network. By being a peer routing node, CEM automatically receives all route updates from other routers in the network and is thus aware of all network events (e.g., network errors, route changes, link failures, convergence) occurring in the IP network.

CEM determines the exact path of the call through the network by correlating session information obtained from the signaling traffic to the IP network topology obtained by peering with the routers. Due to the peering approach CEM has 100% knowledge of the path that a voice, video or realtime data session traverses through the network at any instant in time, both historically and currently, without dependence on NetFlow and JFlow data. If a network event occurs, CEM knows the router interface(s) impacted, and since it also knows the detailed call path, it quickly identifies the specific calls impacted by that network event.

The solution also performs media stream collection and analysis. CEM receives network traffic in the form of RTP packets from a mirrored port (i.e., span port) or passive tap and determines media plane statistics for the sessions being monitored. The solution supports voice and video quality metrics for more than 90 codecs. CEM also captures and analyzes RTCP sender and receiver report information, thereby providing a subset of call quality statistics in the absence of RTP-derived statistics.

What This Means to You

To Customers: The usual home-grown solution for a cloud provider is generally a combination of probes that are installed in the network locally to collect data and isolate problems - SNMP traps, a plethora of other types of point monitoring tools - accompanied by a lot of high end manual effort to diagnose what really happened and at what source. Solutions such as NetSocket's CEM can allow SPs and enterprises to proactively detect, isolate and rectify issues rather than wait until end user problems arise.

NetSocket's Go-to-Market (G2M) strategy targets primarily cloud and managed SPs followed by large (Fortune 1000) enterprises. Today's enterprise sweet spot, however, is in the Fortune 500 realm, while the majority of sales, today, are to SPs.

SPs see NetSocket products as a UCaaS differentiator since it allows them to prove to their enterprise customers that they're delivering service at SLA-committed levels. Furthermore, many of them are looking to resell CEM to their enterprise customers so that enterprise customers can turn around and monitor the services that they're receiving.

On the other side of the coin, interested enterprise customers are typically motivated by the desire to understand the service level that they are receiving from their various cloud SPs. Alternatively, they may be looking to manage the service levels within their own private cloud. And of course they may be looking at doing both - monitoring their own cloud network as well as services being delivered externally and being able to bracket problems so that they know exactly where problems are. In a number of cases enterprise customers have introduced NetSocket to their SP so that E2E solutions can be set up.

In either case, a key differentiator for CEM is its availability both in a traditional Capex purchase model of appliances and software licenses, or in a utility-type model encompassing a complete bundled software suite, all appliances deployed at customer premises, and maintenance and support. Wholesale pricing to SP partners begins at $2/endpoint per month with volume discounts available. This utility-type pricing model leverages the solution architecture and light use of hardware.

The utility pricing alternative and ease of deployment due to lack of network probes and complexity has worked well in greenfields competition against NetScout, NetSocket's closest competitor. None the less the proof of the pudding is always in the eating. As such, we recommend that customers look at all competitive solutions through an RFP solicitation and field trial that aligns total cost of ownership, feature/functionality and development roadmap against business requirements.

To Partners: NetSocket's G2M strategy includes both a direct and indirect sales channel. Its indirect channel is primarily composed of cloud service providers. Direct sales are primarily targeted at large enterprises. Within the channel there are no partners with exclusive rights to sales into any market segment. It's a pretty good bet that the cloud will be carrying more and more of voice, video and data traffic over time. Those SPs looking for an opportunity to ride on that trend have an opportunity to extend their resale product set here.


There are currently no comments on this article.

You must be a registered user to make comments