Got the urge to converge? Wait a while until proven technology arrives.

Another Networld + Interop has come and gone and yet again one of the dominant themes was network convergence.

Not only was convergence the subject of a panel debate involving vendors, analysts and editors, but networking powerhouse Cisco Systems Inc. unveiled a branded line of gear specifically designed to support combined voice/data networks.

While convergence technology has certainly made strides in recent months, enterprise users should be cautious in moving their voice traffic onto their data pipes.

On the WAN side, the benefits of porting voice traffic onto data lines where some sort of QoS can be guaranteed have been realized for several years. If a firm could put up with a marginally lower quality of voice, it could potentially save money by avoiding long-distance toll charges between branch offices.

Stiffer competition in the long distance market, however, is making even this established convergence practice economically questionable.

Even more questionable is the idea of combining voice and data traffic on LAN and campus networks.

Sure, enterprises can save money by paring two networks down to one. One network requires less equipment. It also requires fewer service staff to maintain.

But these cost savings are only worthwhile if the converged network actually works.

Some vendors claim voice/data technology has matured to the point where enterprises can safely deploy converged LAN and campus networks now. Bandwidth on the LAN and campus is cheaper and more scalable than in the past, they say. And emerging QoS tools allow enterprises to give voice packets a higher priority than data packets, enabling voice packets to get to their final destinations with no delay.

Whether the QoS tools work as they’re supposed to remains to be seen. And while bandwidth on the LAN and campus is undoubtedly relatively cheap, putting voice onto a data network requires a highly available, fully redundant network that has the ability to switch from a failed link to a live link in a microsecond. After all, no current data gear provides the resiliency and availability of a “legacy” PBX.

Few enterprises are likely to have this capability in their existing data networks and will need to sink significant amounts of money into upgrading their network hardware and software.

Unless a killer voice/data application comes along in the next few months (there doesn’t seem to be one now), most enterprises would be wise to take a wait and see approach to convergence technology. Some bleeding edge adopters are sure to install convergence gear shortly. Vendors and more conservative enterprises can learn from the early adopters’ mistakes and move to a converged network when the benefits of doing so are a proven fact.

In the meantime, enterprises can keep writing down their PBXs, enjoy the benefits of uninterrupted voice service and shore up their data networks in preparation for the day reliable, proven convergence options arrive.