It didn't take long to see the first signs of strain on communication networks at the London 2012 Olympics when overloaded infrastructure on the first day of competition caused organisers to request that spectators scale back their use of Twitter for "non-urgent" messages, according to Reuters.
On Saturday, during the women's cycling race, broadcasters from BBC had trouble receiving GPS tracking signals from the race course because of an overloaded communications network. Reports then emerged that the number of fans along the race course using Twitter may have contributed to the network strain.
Cisco is the official network equipment provider for London 2012, but the company has denied their involvement: "The tweeting issue had nothing to do with the Olympic network, Cisco or BT," a spokesperson said.
Olympics officials were quoted asking fans to scale back their use of Twitter.
"Of course, if you want to send something, we are not going to say 'Don't, you can't do it,' and we would certainly never prevent people," an International Olympic Committee (IOC) spokesperson was quoted as saying in Reuters. "It's just - if it's not an urgent, urgent one, please kind of take it easy."
Reuters reported that during the situation on Saturday, "many inadvertently made matters worse by venting their anger on Twitter at the lack of information."
How could this happen? Brian Jacobs, senior product manager for the network management division of Ipswitch, said the situation highlights the need for organisers of large events to have networking monitoring and controlling tools in place.
"Whoever set up this network obviously didn't fully think through the impact of the fans and their ability to consume a massive amount of bandwidth," Jacobs said.
Given all of the preparation Olympic organisations have done to set up the networks to carry the games, he said there are lessons enterprise network managers can glean from the situation.
First, he said, it is key to have network monitoring tools in place to gain insight into the capacity of the network and the traffic on it. It's one thing to know that the network is becoming overloaded, but it's another to know what sort of traffic is causing the issue. "Gross-level" monitoring will track the overall capacity of the network being used while fine-level monitoring will show exactly which traffic is causing the stress.
The next step is having tools in place to manage that network load. Protocols can be put in place to prioritise network traffic and limit some sources that are requesting bandwidth. A variety of players, including Blue Coat Systems, do that sort of work, Jacobs noted.
"You can't treat all the traffic the same," he said. "Business critical apps need to take priority over nominal communications and there are ways to configure the network to do that."
And finally, he said, a lesson from the Olympics issue is that you can't blindly rely on your partners. The issue over the weekend, he noted, was likely caused not only by the Olympics network infrastructure having issues, but also from third-party telecommunications systems that may have been overloaded. If an enterprise is relying on a partner or vendor to supply a networking service, make sure the provider is putting controls into place to handle unexpected issues that may arise as well.