Mea Culpa: Time To Build Security Into Connectivity
How those of us who spent decades developing faster, easier, and more scalable networking technology have made the lives of our security counterparts a living hell.
I have been very lucky in my career. I started in Chemical Engineering but the labs were too smelly and dangerous for me. So, I switched to the relatively clean world of Electrical Engineering in my senior year. How’s that for career decision-making criteria? But, it worked out. At some point I saw the value of Ethernet and TCP/IP to foster scale and ubiquity. I have been focused there pretty much ever since.
I am not unique. A whole generation of engineers, architects, and marketing people have been evolving packet networking technology over the past three decades. I, along with many others, have been involved in the creation and promotion of several generations of speed improvements, cost effective expansion of connectivity to the home, “untethered” wireless and cellular connectivity, bandwidth optimization methods, clustering techniques to match the speed of networks to servers, and as a foundation for all of this – DNS, the “reverse white pages” of the Internet that allows you to learn anybody’s address from their name. We’ve spent our lives continually enhancing the ease of access to, automation, and scale of connectivity.
Good job, everybody!
But there is a bit of a downside to this success story. While those of us creating ever easier and scalable connectivity have had some fun times, we have made the lives of our counterparts in the security industry segment a living hell.
All that easy connectivity starts with visibility: server addresses advertised to anyone who asks, servers continually listening to and responding to anyone who wants to connect. The same foundation that has created the ever expanding Internet has provided an easy path for any attacker to exploit and compromise or deny service to any application running on the Internet. Easy connectivity also ensures that a large number of users are exposed to and infected by that exploit, probably then passing it along to other users and applications. The beat goes on.
The result is that our security brethren have to play a constant game of “whack-a-mole” to discover attacks in the noise of all the legitimate traffic that is running on their systems; usually discovering and remediating long after the attackers have done their dirty deeds. It’s a losing battle. Even if they find a useful security control, someone invents a new way of achieving connectivity, or a new economic advantage that perturbs the IT deployment architecture, and the new control rapidly becomes yesterday’s news while embarrassed compromises become today’s news.
Spending in security has risen from about $45B per year to about $90B per year in a few short years, but I challenge you to find someone who feels they are more secure today than they were a few years ago.
Houston – we have a big fundamental problem here.
I think that those of us who have created easier and more scalable connectivity have always relinquished the job of security to others. And those others are doomed.
Security is not a “finishing touch.” You can’t build the IT framework without security and then achieve security through spackle and a fresh paint job. Security is not a “punch list.” Security is not an afterthought, an accoutrement, a garnish, a Kardashian, a veneer, or a side dish.
Security needs to be built into the very foundation of connectivity rather than left as a homework exercise for the reader (i.e. poor security schmuck). The world needs connectivity to be intrinsically secure.That won’t solve every security issue in the world, but it sure could establish a workable foundation for other security tools to be effective.
It is time for those of us who created these problems to help solve them. This is why in the twilight years of my career I have now become a “security guy.” That mostly means I have had to learn to say words such as “mitigate,” “elliptic curve,” and “exfiltration,” without giggling. But it also means I’ve had to look for a model, framework, architecture, or whatever you want to call it, to build security intrinsically into connectivity.
But the purpose of this blog is not to tout what we are doing. The purpose of this blog to urge all clever people working on networking to think of additional ways to build security into their solutions so that IT security people can occasionally go to a Sunday cookout and relax a little. It’s a noble cause. And for you Silicon Valley people, let me speak more in your language – there is a mountain of money to be made by anybody that can help get this right.