|
11/2001 - UCI (Unified Comms International) magazine - Rising from the Ashes
|
The cables that run beneath the streets of Manhattan buzzed with
traffic after 11th September. So how did the networks cope, and what
lessons can be learned?
|
| |
The recent terrorist attacks in the US horrified the world, and left
the financial hub of Manhattan paralysed. Newspapers were full of tales
of “meltdown” and “collapse” in the networks. However, on closer
reflection it emerges that many of the communications networks were
surprisingly robust in the face of such damage. Manhattan is by far the
largest hub for long-haul Internet bandwidth, connecting to 102 cities
in 71 other countries in the world. The four largest inter-regional
routes in the world link to New York, including the New York to London
route, which, at nearly 78Gbps and almost quadruple the next largest
route, forms the international Internet’s main geographic trunk
(statistics from Telegeography’s upcoming report Packet Geography 2002:
Global Internet Statistics & Commentary).
The distributed nature of networks means that if one switch goes down,
traffic can be re-routed. Also, as carriers regularly rent space from
each other, the processes are in place to re-establish service in the
event of any local problems. Capacity is not an issue here – the
problems many of us experienced when trying to call the States on the
11 September were caused by the destruction of switching equipment in
the financial district of Manhattan itself.
On the actual day of the attacks, long-distance carrier AT&T
reported seeing a ‘spike’ in voice traffic towards the East coast of
the US, primarily New York and Washington, and had received about 600
million calls by noon EST (three hours after the attack), which is
twice the normal daily volume. According to Frank Ianna, CEO, AT&T,
the network (voice, data, IP, frame and ATM) “held up quite nicely”.
All long-distance capabilities were intact, but as a local node in the
basement of the World Trade Centre had been destroyed, local voice
switching capacity into some areas of Manhattan was lost. AT&T’s
emergency procedures include trailers which can roll capacity into any
affected area from across the country, and after giving all possible
help to the emergency services, the strategy was to initially get
connectivity to the customers in the downtown area, using Verizon’s and
AT&T’s cables, and especially to get the Stock Exchange ready to
function. The next priority was to get more capacity into the region so
that companies unable to get into their own buildings could find new
locations.
New York-based carrier Verizon had a network switching centre on the
tenth floor of one of the collapsed World Trade Centre towers, as well
as a facility at West Street which housed more than 1,700 employees and
which affected 200,000 residential and business telephone lines, and
more than three million data circuits. Verizon said they have had 1,600
employees working around the clock since the attacks to restore service
to the New York Stock Exchange and the major brokerage firms, many of
which had relocated to new or temporary offices. By the time the NYSE
reopened on 17 September, service was restored. “The NYSE capacity is
at almost 100 percent. At even less than 100 percent, we’re able to
handle volume that is significantly in excess of normal demand,” says
Larry Babbio, vice chairman, Verizon. On its first day of opening after
the attacks, the NYSE handled its heaviest trading volume ever, with
more than 2.3 billion shares changing hands.
It was not just fixed line networks that were affected. Wireless
operator Nextel reported “extraordinary caller volume” along the East
Coast, and a sharp upsurge nationally, and said that the most effective
method of communication for its customers was two-way radio or text
messaging, neither of which rely on the public switched network. Since
the attacks Nextel has loaned over 5,000 phones with Direct Connect
capabilities to federal, state and local government agencies, and plans
to co-operate with Motorola to provide up to 12,000. Sprint, AT&T,
Verizon and foreign carriers have all been generous and quick to
provide prepaid calling cards, wireless handsets and free calls from
payphones in the affected areas. Michael Powell, chairman of the
Federal Communications Commission (FCC) has nothing but praise for the
co-operation shown between companies, “I can report how deeply
impressed and proud I was by both the intensity of the repair efforts
and the commitment that the industry is bringing to this task. These
companies know how important telecommunications is in our society
today, and how essential it is for consumers and businesses alike to
have their damaged or destroyed lines operational as soon as possible.”
He describes Verizon’s efforts as ‘Herculean’ and pledges his support
for the communications companies in the weeks to come.
Frustration
All of us who were trying to contact friends and family in New York can
testify to the frustration of not being able to get through, and voice
services were the worst affected. BT reported that from the UK, there
was a dramatic increase in transatlantic calls. In a normal 15-minute
period, one switch takes around 700 calls to the East Coast of the US.
At the peak of the telephone rush at around 3.15pm (GMT) this number
rose to 60,499, an 86-fold increase. As a result of the surge in
demand, call gapping (restrictive controls at local exchanges) was
implemented to protect the network and ensure that at least a
percentage of calls could get through. However, many network operators
report that the problems were not in the backbone of the network, but
rather in the local access. Essentially, if cables and switching
facilities in Manhattan have been destroyed, no amount of capacity can
get calls through.
This was a time when other channels of communication really came into
their own. Although servers for news sites such as CNN and the BBC were
swamped with demand, and there was a bottleneck of traffic into the
Manhattan and Washington areas, other networks, for example, BT’s IP
network, were completely unscathed. In terms of web presence, the main
news sites suffered because their ISPs couldn’t scale up to the huge
demand surge. There are queries as to what the TV networks should do
about this, as the cost of adding more capacity to deal with any kind
of demand shock could be prohibitively expensive. Again, these are
problems at the local server level rather than larger network issues.
Jessica Marantz, business development director, Telegeography points
out that the Internet is so robust because it was initially designed to
keep government going in times of war, and it was generally seen to
have held up better than voice traffic, especially mobile. LINX (the
London Internet Exchange), Europe’s largest Internet hub, reported a
ten percent dip in Internet traffic on the day of the attacks,
presumably when a number of companies were sent home, and as people
turned to TV for information, and then an increase over the subsequent
days.
Matrix.net, a provider of Internet performance measurement products,
reported just hours after the attacks that, “the main Internet
infrastructure in the continental US appears to be functioning close to
normal levels.” Shortly after the attacks, Matrix.net detected a
significant performance degradation of major Internet infrastructure
points and Web sites (as measured by increased packet loss and
decreased reachability). While the spike was significant, it was
relatively short-lived, and IP traffic returned to near-normal
performance levels within about an hour. A spokesman for Matrix.net
continued, “The Internet was designed to withstand stresses, and it is
currently functioning well under the greatest stress encountered over
its 32-year history.” Whatever concerns may be harboured about the Web
as an instrument for terrorists to use to communicate, in this first
great test, the Internet as a method of communication for the worried
and the bereaved came out with flying colours.
Virtual Options
Now that the relief effort is well under way, broader questions are
being raised about what implications these tragedies will have on the
way we communicate, in business and in leisure. Could this be the time
that video-conferencing and audio-conferencing eventually reach a mass
market? Business travel has been severely curtailed by security worries
and the possibility of war, and companies are increasingly looking to
alternative means of contact with internal colleagues and with clients.
Shares in the world’s second largest video-conferencing equipment
manufacturer, Norway’s Tandberg, have risen more than 25 percent since
the disaster.
Although events such as the fuel crisis and train crashes in the UK
led, in the past, to an increase in demand for these facilities,
video-conferencing has never quite lived up to its promise as a
business application. David Sales, director, BT Conferencing feels this
may be about to change. On Wednesday 12 September, the day after the
attack, BT Conference saw an increase in demand of 70 percent. This
levelled out to 30 percent over the subsequent days, and he expects
this to continue. Glenn Manoff, VP corporate and public affairs, Ebone,
comments, “My view is that web-based video-conferencing is an effective
option for many and will be affected more than traditional
video-conferencing.” Video-conferencing manufacturer Polycom have
offered free use of their facilities until the end of September for
those affected by the attacks, whether they are displaced companies or
families unable to meet up face-to-face. Cognigen Networks will also be
offering a new teleconference service aimed at social groups and
families. The facility has traditionally been used only by businesses,
but Darrell H. Hughes, CEO, Cognigen, thinks recent events will change
that.
“One certainty is that Americans, both at a leisure and business level,
will communicate differently, with less face-to-face and more
tele-connected communication.”
Remote access
This should surely have been a time when the benefits of outsourcing IT
and data functions would be demonstrated. Blue Ridge Networks, a VPN
solution provider, has donated remote access VPN service to companies
that were displaced or who had their infrastructure damaged in the
attacks, but short-term solutions can only achieve so much. Companies
will need to plan for the possibility of future attacks, and colocation
facilities may have to rethink their service offerings in light of
these attacks.
Telehouse was one of a number of companies who suffered physical damage
in the attacks. The Telehouse facility at 25 Broadway in Manhattan
suffered a generator malfunction, and power was not completely restored
until three days later. Analysts are expecting managed data services
companies to do well in the coming months, as businesses may find it
more effective to outsource IT departments and thus avoid the headache
of protecting their systems and data. Vanessa Evans of LINX believes
that as demand goes up, companies will now also be demanding more from
their colocation facilities, including secret locations and heavy-duty
protection against natural and manmade disasters. The last thing anyone
wants is for these facilities to become a viable target for groups
looking to cripple the economy. “Security, which was already one of the
most important issues, will obviously become even more important. The
key elements are data security and network security,” comments Glenn
Mannoff of Ebone.
Disaster Recovery
Frank Ianna describes AT&T’s disaster plans as being very
successful, but adds that they had never planned for a situation where
it was not possible to fly key personnel to the affected locations. No
one had envisaged a situation as extreme as this, but precautions that
companies can take include using broadband connectivity to provide
alternative communication channels in an emergency. Experts also advise
having corporate cellphones on a number of different networks,
providing a back-up website in a separate location, and backing up key
servers and files regularly.
Disaster recovery experts Comdisco report that at the time of going to
press, 35 of its customers (including retail, investment and foreign
banks, insurance providers and media companies) have declared disasters
requiring the invocation of disaster recovery plans, and 13 out of its
45 recovery centres have been utilised. “The other critical thing
companies need to remember,” says Raj Goel, CTO, Brainlink
International, “is that not all disasters are the same. Companies
should not be focusing solely on the possibility of another terrorist
event. It is important to remember that in this changed environment
companies are still vulnerable to natural disasters, significant
equipment failures, power outages and all the other types of
comparatively mundane problems that disaster recovery planning was
originally designed for.”
The one thing that has struck us in the world of telecoms is that
however many new technologies we can invent, telecoms is still at the
end of the day about communication, and it still has a crucial role in
society. Those haunting last mobile phone calls from hijacked planes
were testament to our basic human need to communicate and the
importance of these technologies to us now. In the profit-driven world
of deregulation, it is heartening to know that the telecoms companies
have not forgotten their key role as critical public service providers.
There have been many tales of heroism and of employees going well
beyond the call of duty to ensure connectivity for their customers, and
we certainly salute them for it. Our thoughts and prayers go out to
those who are grieving now. |
|
Press Coverage