Incremental Deployments

As previously mentioned, convergence is going to challenge the traditional thinking that goes on inside corporations. How companies deploy technologies, the expectations they have for voice technologies, and how they ultimately support these new technologies is undergoing dramatic change. This makes many people uncomfortable because of the misconceptions surrounding IP telephony.

Part of the change begins with how this technology can be deployed. Clearly, one of the major initial benefits of IPT is that it can be deployed incrementally. Corporations have the opportunity to protect (and if desired, expand) the investment already made in TDM technology while at the same time beginning to deploy IP telephony in a cost-effective manner. This approach is similar to what happened in the 1980's migration from centralized mainframes to LAN-based solutions, as shown in Figure 4-6.

Figure 4-6. Migration from Mainframes to Distributed LANs

Figure 4-6. Migration from Mainframes to Distributed LANs








Open Distributed Computing

Think back to when organizations first began migrating from mainframe computers to LANs: No one threw their mainframe out the front door right away. Indeed, many companies today still have both mainframe computers and LANs in place. Organizations, department by department, migrated (or transitioned) their users from being mainframe-connected to becoming LAN-connected. After users were connected to the LAN with a new "workstation" (the personal computer), they still had to maintain connectivity for that workstation back to the original mainframe. This enabled users to access their existing applications, while taking advantage of new ones that were more departmental, or even local to their machine.

As new applications became available, end users began to depend more on their local PC and departmental servers, and less on the traditional applications on the mainframe.

Subsequently, many of the mainframe-based applications were migrated to the LAN environment with enhancements. In the end, some companies eventually did away with their mainframes, and others still use both mainframes and LAN applications to serve their employees and customers.

The methodology for migrating from one data environment to another was demonstrated in clear terms in the 1980s and 1990s. Now, much of the same thing is happening. The mainframe has been replaced by the PBX. The distributed LAN is replaced by distributed IP telephony. Yet, the methodology that was proven in the earlier migration remains the same:

• In the 1980s, the older 3270 terminals were replaced by personal computers. Today, older PBX phones are being replaced by IP phone clients.

® In the 1980s, personal computers were taken off the mainframe (with their 3270 emulator cards) and placed on the LAN. Today, new IP phones are being placed directly on the LAN (IP network).

® Finally, in the 1980s, LAN workstations were provided access back to the mainframe environment and its applications by means of a gateway. Today, a new kind of gateway provides connectivity for LAN-based telephony users to access resources back on the PBX.

The technical merits, challenges, and solutions of migration strategies are explored in Chapter 7, "Watch That First Step"; however, it bears noting that this technology was designed, from its inception, to be migrated into a customer environment following the methodology learned from the mainframe-LAN transition.

Just as the expectation was not to completely remove one's mainframe computer immediately, the expectation for IP telephony is that companies will not want to replace their PBXs completely all at once—because they don't have to.

This sounds fairly simplistic, but the truth of the matter is that this has never happened before with voice communications. The procedure has always been to buy a PBX and flash cut it over a weekend, or overnight. The financial investment was a replacement investment; that is, companies had to invest enough to replace the entire system.

Now, however, IP phones can be deployed into an organization individually, or departmentally, as is the case with other IP clients, such as PCs or printers. IP phones can be implemented on the IP network and allowed to coexist with the traditional PBX phones, just as was done with PCs and mainframes. This means that a company can begin deploying a new phone system—a new telecommunications solution—into their organization a piece at a time. This is something that has not been cost-effectively done with traditional PBXs. Instead of a replacement investment, IP telephony enables companies to make transition investments; for example, investing only in the technology required to migrate x number of users, where x is considerably less than the entire organization, and is determined by the customer.

This manageable and measurable approach can turn the stressful event of cutting hundreds or thousands of phones and users into an easily managed process that can take as little or as much time as the organization decides. This approach has become a time-tested methodology within the IP telephony market in the past few years.

Consider the fact that Cisco Systems took almost 3 years to migrate all 23,000-plus of its employees in its San Jose campus environment away from its legacy PBX to IP telephony. In January 2003, Cisco cut over the final piece of TDM switching from their Lucent Definity system to the Cisco AVVID IP telephony solution. This project began in 1999, and it followed the incremental migration model. Cisco went department by department, building by building, learning from each deployment and adjusting accordingly. For Cisco Systems, the potentially daunting task of moving from TDM to IP became an easily managed process.

This is the new model for deploying telephony. For a period of time, two different environments are supported instead of one. Usually, one of those environments is new for the people providing support. The telecom staff must become familiar with IP, and the IS staff must become familiar with voice.

Subsequently, that familiarity must lead to true competency. An incremental deployment, or migration, allows this learning inside the organization to happen gradually, at a pace comfortable for the personnel involved. Everything becomes incremental: learning, spending, training, and ultimately, deployments.

This methodology also provides a tremendous cultural benefit to the organization. It allows the IS and telecom groups to learn to work together, and clearly shows each group that one is not a threat to the other. Both groups remain essential to the success of the organization.

Was this article helpful?

0 0

Post a comment