September 3rd; 1967 was "Dagen H" (H day), or "Högertrafikomläggningen" ("The right-hand traffic diversion") in Sweden. Car traffic switched from driving on the left-hand side of the road to the right. All non-essential traffic was banned from the roads from 01:00 to 06:00 and the remaining traffic had to follow special rules. Vehicles had to come to a complete stop at 04:50, then carefully change to the right-hand side of the road and stop again before being allowed to proceed at 05:00. Larger Stockholm and Malmö saw an extended ban to allow work crews to reconfigure intersections. Bus stops had to be constructed on the other side of the street. Intersections were reshaped to allow traffic to merge. Over one thousand new buses were purchased with doors on the right-hand side. Some 8,000 older buses were retrofitted to provide doors on both sides, while other buses were sold to Pakistan and Kenya where the traffic remained on left-hand side. Swedish Dagen H is not the only example of massive traffic diversion: Czechoslovakia switched to right-hand traffic on May 1st; 1939. On "730" (Nana-San-Maru) - July 30, 1978 - Okinawa Prefecture of Japan switched back from driving on the right-hand side to the left. On May 26th; 1968 the "H-dagurinn" (Right day) saw Iceland changing from left hand to right hand traffic. Samoa moved to left-hand traffic in September 2009…
In the enterprise IT world, Dagen H situations (switching from one business application, a datacenter or an operating system to another one on a single day across the entire organization) are not an option, as they would raise the cost and the level or risk for projects – not speaking about the requirement to keep the business flowing in every circumstance. According to the size, the organization and operational constraints within organizations, significant IT transformations like generalizing Windows 7, moving from Notes to MS Office, from Exchange to Office 365 or GoogleApp, from PABX to Unified Communication, from PC to light terminals, etc. take from several quarters to a few years. They are always complicated projects involving many aspects, the technical ones being often not the most difficult: ignoring employees' culture, underestimating resistance to change or missing the opportunity to present the benefits to users turned more than one project with perfect technical management to IT Black Swans.
These long periods where "before" and "after" must live together are critical as managing transitions is always tricky. Whilst the behavior, limitations and performance of the old system are well known, the adoption of the new one is quite difficult to anticipate as usages will stabilize only after a significant amount of time. Who can really forecast the real usage of desktop video within an enterprise after the deployment of its Unified Communication system? Will users rapidly limit its usage to a few well defined situations or will "one-click video conferencing" become the norm? Will Open Office or the new Google suite be as effective and compatible enough with MS Office used by employees, partners and customers? Will MS Lync voice and video be properly handled through the new desktop virtualization protocol in all cases? Will streaming news and attend on-line training become mainstream within the entire company or remain a nice to have?
Uncertainty is certain. IT organizations must put in place adaptive processes and tools that can handle the unknown in a reasonable and fast manner. As most of the new usages converge to the corporate (private or hybrid) network, IT performance depends largely on how applications flow over it. Controlling these applications over the WAN is a task more difficult – and more critical - than ever. Addressing the challenges above is the raison d'être of network Application Performance Guarantee, WAN Governance and Autonomic Networking.
Illustration: "Dagen H" in Stockholm - September 3rd; 1967 and the "Dagen H" logo.