It doesn’t take a crystal ball to prepare for future threats to corporate network security. All it takes is a library card.From the larcenous potential of remote-control crime to the progressive obsolescence of every known security architecture, science-fiction writers have anticipated both the social factors and the big-picture technical developments that have shaped and directed the network manager’s agenda. For at least two decades, high-tech adventure novelists and avant-garde “cyberpunk” renegades have demonstrated remarkable foresight.
It may be easier to get excited about the coming challenges by reading a believable story than by facing the daunting prose of a comprehensive reference on this subject. Nonetheless, Rita Summers’ 1997 “Secure Computing: Threats and Safeguards” (McGraw-Hill, ISBN 0-07-069419-2) is an excellent source of information.
An IBM veteran, Summers gives the network manager a framework on which to build a site-specific plan, beginning with the fundamentals: accountability, awareness, ethics, proportionality of protection vs. risk and multidisciplinary design with planned reassessments to deal with inevitable change in the system’s environment.
Summers’ 360-degree view is essential. Simple but comprehensive systems, implemented well, will provide more protection than more complex systems if the costs of the latter lead to gaps in coverage of secondary risks, or if a sophisticated design invites small but crucial configuration errors.
It’s all too easy to get wrapped up in the furor over exotic, newsworthy developments in disciplines such as cryptography, while ignoring the mundane reality that many users choose their mothers’ maiden names as their passwords. A network can use a kilobit key and a state-of-the-art algorithm to protect data confidentiality and integrity, but with careless or uneducated users, the result is no more secure than a massive bank vault with the combination written on the door.
Perhaps even this analogy is too kind. The bank vault, at least, cannot itself be stolen, but many networks are both easy to crack and easy to crash. The former deprives the network’s users of intangible benefits such as privacy; the latter deprives them of tangible business opportunities that come and go while a network is down.
Crashing a network is all too easy.
It’s meaningless, for example, to obsess over the crash protections afforded by redundant servers with failover operating systems if both of the servers share a single source of electrical power.
It’s missing the point to have RAID-based, hot-swappable disk arrays in a room that would be flooded by water from the engine company’s hoses in the event of a building fire.
Less obvious, but increasingly likely, is the threat from electromagnetic energy, whether accidental or malicious. A microscopic chip, whose millions of transistors each dissipate only microwatts of power, can be fatally damaged by an all too easily generated beam of microwaves.
The good news is the spreading awareness that security must be designed in, not added on. Important emerging technologies such as Java and IPv6, for example, included provisions for secure computing in their original definitions of the problems that they were created to solve.
The bad news is the degree to which law, employment practices and other social frameworks are not staying abreast of the soaring value of electronic transaction flows and centralized collections of data. The ubiquity of the Internet makes it cheap, fast and easy to integrate many different sources of information in ways that threaten privacy, despite the seeming innocence of each individual request for personal data.
The growing convenience and transparency of wireless connections make it easy to overlook the relative simplicity of plucking data, literally, out of thin air, without committing any physical and easily detectable crime such as trespass.
At least technology is finally beginning to help the good guys as much as it does the malefactors. The same fast microprocessors that enable brute-force attacks on encryption schemes also enable automated recognition of voice, face, fingerprint and iris patterns that establish personal identity with high accuracy and minimal inconvenience.
High-speed data links make it practical to incur the overhead of wrapping security headers and other protective bit groups around valuable data. Packet-sniffing tools, already used to diagnose low-level problems in network operation, can be applied at more refined levels to highlight suspicious activity.
With CPU cycles to spare, network clients and servers can do the formidable computations involved in knowledge-free key exchanges and other mathematically intensive cryptographic protocols while barely disrupting normal operations (as long as appropriate methods are used to prevent an attacker from bogging down a server with spurious key requests).
But network managers can never become complacent about any protective mechanism. It’s easy to deceive one’s self into thinking that a security architecture has multiple layers, when it really just has multiple links with the whole as weak as its least secure component.
For example, a manager who thinks that face or voice recognition is inherently safer than passwords is overlooking the fact that both of these mechanisms must reduce, at some point in the process, to the right set of bits crossing the right connection at the proper time.
A determined network cracker who’s willing to take the time, and spend the money, required to sniff the right packets can forge a face–electronically, that is–as readily as he can sniff an unencrypted password. More distributed networks create ever more opportunities for the physical access that enables an electronic attack.
There is growing recognition of the inexorable progress of cracking tools and techniques and of the inevitably statistical nature of all protection.
A lively, constructively paranoid imagination is therefore, perhaps, more important than any other single intellectual asset in defending corporate networks against the threats to come. Few things make less sense than spending money and time installing and administering security technologies and policies before there has even been any systematic effort to identify and estimate both present and foreseeable risks.
It’s important to think in terms of long-term trends. This may explain why the predictions of cyberfiction have been so remarkably prescient, despite many writers’ non-IT backgrounds.
Vernor Vinge’s 100-page story, “True Names,” envisions the fusion of gaming techniques with worldwide communications: Vinge has suggested that his ignorance of computer technology (at the time he wrote the story) might actually have been an advantage, “saving me from getting lost in the irrelevancies of the moment.”
That comment appeared in Vinge’s foreword to that 1981 story, as republished in a 1987 collection by the same name (Baen Books, ISBN 0-671-65363-6). “After all,” Vinge continued, “I knew where things were ultimately going!”
William Gibson wrote his 1982 short story, “Burning Chrome” (Ace, ISBN 0-441-08934-8), which coined the word “cyberspace,” on a manual typewriter. Ditto his 1984 novel “Neuromancer,” another of this genre’s defining works (Ace, ISBN 0-441-56959-5). Gibson adopted a computer only when his trusty machine broke a critical, out-of-production part, and he was told that a new equivalent model would cost more than an Apple II.
In this respect, Gibson demonstrated a principle that network managers would do well to follow. He did not assume that the latest technology was by definition the best; rather, he waited until the cost of continued maintenance exceeded the cost of upgrade. The function and the cost, not the technology, determined the timing of his moves.
Effective network security policies are no different. Focusing on the functions of the network, rather than its technologies, and on the relative magnitudes of risks and remedies, will minimize waste of time and money while affording the protection that information assets deserve–now, and in the decades to come.
“It was a typical banking defense and cash-flow plot–that is, typical for the SIG. Most banks had no such clever ways of visualizing the automated protection of their assets. …
“He poked a finger into the map and a trace gleamed red through the maze. ‘If they’re lucky, they’ll discover this tap next autumn, when they find themselves maybe three billion dollars short, and not a single sign of where it all disappeared to.’ ”
–”True Names,” by Vernor Vinge, Baen Books, New York, 1981
A visualization tool called HyperSpace (www.cs.bham. ac.uk/~amw/hyperspace/www95) creates analogies between familiar ideas (like size and distance) and abstract concepts (like download time or resource accessibility).
Other, more concrete visualizations of network activity are online at www.nlanr.net, a site maintained by the National Laboratory for Applied Network Research.
Related article: PC Week Recommends
Investigate tools for visualizing network configuration, activity and applications performance that may accelerate identification and analysis of attacks.
“The intrusion countermeasures had accepted his entry as a routine transfer from the Los Angeles complex. Behind him, viral sub-programs peeled off, ready to deflect the real Los Angeles data when it arrived.”
–”Neuromancer,” by William Gibson, Ace Books, New York, 1984
“One of the vandals’ favorite sports was to infiltrate office systems and simulate higher level input to make absurd and impossible demands on the local staff.”
–”True Names,” by Vernor Vinge, Baen Books, New York, 1981
“The IPv6 specifications [describe] an Authentication Header and an Encrypted Security Payload. One provides authentication, guaranteeing that the source address is authentic and that the packet has not been altered during transmission. The other guarantees that only legitimate receivers will be able to read the content of the packet.”
–”IPv6: The New Internet Protocol, 2nd Edition,” by Christian Huitema, Prentice-Hall PTR, New Jersey, 1998
Related article: PC Week Recommends
Establish a plan for retrofitting IPv6 procedures to IPv4 systems and/or make the transition to IPv6 throughout all critical networks.
“Look, this stuff is legal. That gadget is scarcely more powerful than an ordinary games interface.”
–”True Names,” by Vernor Vinge, Baen Books, New York, 1981
“Even amateurs have access to the technological tools needed to penetrate systems and cause trouble.”
–Robert T. Marsh, chairman, President’s Commission on Critical Infrastructure Protection, January 1997
A good drug rehab counselor must be determined to help his or her patients. They must have the spirit to walk with the patients through the whole program helping them to overcome their addiction problem. Your first visit to the rehab center will help you to know whether the counselor is fit for your patient or not. Secondly, a good drug rehab counselor must portray a great sense of patience. It is not easy to work with addicts. Some of them become rowdy and difficult to handle especially on their first days at the rehabilitation centers. A good counselor will be ready to handle them in this state with a sheer determination of helping the patients to overcome their craving.
