Events Training Consulting Newsletters Webcasts Blogs
Subscriptions
Current Issue
Past Issues
Join Our Mailing List
Contact Us
Home
 
 
 

 


TechEncyclopedia

Virtual Private Networking Version 2.0

Technological and market developments are pushing VPNs into the next evolutionary phase. What forces are shaping this second iteration?

By Elizabeth Clark

print this article print this article
email this article e-mail this article
.

Cox Communications
Cultivate An On-Demand Workforce Through On-Demand Technology
CallCopy and SER Sign Partnership Agreement
Telrex Announces Support for Cisco System
The Fading Away of the TUI
Q and A: The Importance of Testing Your Technology
Q&A: Interactive Intelligence Helps Migrate to Total UC Faster
When Push Comes to Shove
CSI Arms Virtual Observer With New Toolset
Quality & Recording Solutions Keep Getting Better
.

11/01/1999, 12:00 AM ET

Much like version 1.0 software, the first wave of VPN technology suffered its share of bugs and blunders. For some users, VPN setup hasn’t been the carefree experience the system vendor described. Many IT staffers have felt the fallout of skimpy management features. Others who opted for managed services have been cast adrift in a sea of Service Level Agreements (SLAs), security features, and connectivity options they don’t fully understand. And then there are those who’ve endured fidgety meetings with the CFO, trying to explain why the organization’s VPN hasn’t saved quite as much money as originally projected.

Compounding the situation has been the disproportionate hype that has swirled about in the VPN market. In reality, neither user adoption rates nor cost savings have lived up to initial forecasts.

But VPN technology, as well as the market forces affecting its development, is changing. Recent advances in security and QoS, new outsourcing opportunities, and the rising number of carriers and service providers offering VPN solutions are just some of the forces shaping this evolution.

As interest in VPNs has increased, so have the number of VPN services. VPNs are now being used for remote access, site-to-site connectivity, extranets, and intranets. They’re also being used to secure internal traffic within individual organizations—in effect, supplanting VLANs.

A study conducted by TeleChoice (www.telechoice.com), a telecommunications consulting firm, shows how adoption trends and users’ concerns are evolving. According to its “TeleChoice VPN Market Report,” which surveyed 501 organizations, 130 companies indicated they already have a VPN, while 175 said they are planning to implement the technology in the next 18 months.

One factor affecting the adoption rate is user priorities. According to the TeleChoice report, performance, security, ease of use, and price (in order of ranking by respondents) are the top criteria for choosing to adopt a VPN. While progress has been made in many of these areas, work still needs to be done to overcome some of the barriers that have made organizations reluctant to use the technology.

Another factor affecting VPN adoption rates is the option to outsource remote access. According to a recent study by Forrester Research (www.forrester.com ), this is the primary driver behind users’ growing interest in VPN technology (see Figure 1 ).

The report also indicates that 46 percent of survey respondents plan to build their own VPNs, while 33 percent will opt for a managed service. Figure 2 shows why some respondents would rather keep VPNs in-house. (For more on outsourcing, see External Affairs.)

When considering a VPN implementation, a major issue for potential users is whether to connect over a private network or a public IP-based VPN. While numerous options are available on both sides of this fence, organizations need to carefully evaluate the properties of each alternative.

On the public IP network side, QoS across the Internet is not yet a reality. In addition, these networks raise security concerns among some potential users. On the up side, Internet-based VPNs are widely available and relatively economical.

VPNs delivered over private networks, such as frame relay or ATM, provide a level of QoS control that can be difficult to duplicate over the Internet. However, these configurations can be expensive and don’t necessarily offer more solid security features than their Internet-based counterparts.

Fortunately, additional options are available on both the public- and private-network side, which is helpful for organizations trying to match their data communications needs with the proper medium.

PASSING THE PERFORMANCE TEST

Many recent developments in VPN technology have centered around performance issues, particularly QoS. Both service providers and vendors are scrambling to address this major concern.

The customer’s primary tool in effecting QoS is the SLA. User demand for such agreements is growing, and providers and carriers are responding with more stringent SLAs. For example, Concentric Network’s (www.concentric.com ) ConcentricQoS service includes a 100 percent, premises-to-premises network availability guarantee and a maximum latency guarantee of 80 milliseconds (ms) for VPN connections.

Concentric’s guarantees are based on performance as measured by Inverse Network Technology (recently purchased by Visual Networks), which provides service-level management software and Internet benchmark services.

For remote-access VPNs, the SLA stipulates a busy-free dial availability of 97 percent, or less than the industry average (as measured by Inverse Network Technology). The latency guarantee specifies an initial modem connect speed of 26.4Kbits/sec (minimum) at 99 percent, or less than the industry average (also as measured by Inverse Network Technology).

Living up to SLAs means overcoming some significant hurdles. “QoS is challenging from a carrier perspective because we’re now becoming even more involved with the LAN,” says John Lawler, VPN product manager at Concentric Network. “It used to be very simple to be a carrier—we stopped at the router. But now we’re going further and further into the customer’s network.”

At the heart of the QoS issue is effective traffic prioritization. Current mechanisms include Multiprotocol Label Switching (MPLS), Differentiated Services (DiffServ), Class-Based Queuing (CBQ), and Common Open Policy Service (COPS).

MPLS, which is based on an IETF spec, defines a process in which a label is attached to an IP header to increase routing efficiency and enable routers to forward packets according to specified QoS levels. With MPLS, routers can assign explicit paths to various classes of traffic.

MPLS uses a tunneling technique in which IP traffic is encapsulated in new routing headers. An MPLS-enabled edge router adds a label to the packet at the edge of the network. Forwarding is based on information in the fields of the label that has been added to the packet. As the packet travels through the network, it may be relabeled numerous times. The label is removed when the packet reaches its destination. (For more details on how MPLS works, see “MPLS: A Progress Report.” )


| 1 | 2 | 3 | 4 | 5 | Next Page > >

.

Free CallCenter Insider Newsletter

Your Email Address


Optional Areas of Interest
International News
Advice/Tips
Technology
Agent Development
IVR