It’s worth noting that even sleep-deprived Rich is surprisingly coherent.


While the RSA show technically doesn’t start until tomorrow, there’s still a heck of a lot going on. For myself, the worst is actually over. And by “the worst”, I mean there are even odds I will actually sleep tonight.

It all started yesterday when we delivered the very first CCSK certification class for the Cloud Security Alliance. I learned three things in the process:

  • Managing other analysts on a project sucks major @$$.
  • We totally need 2 days to cover this content. Heck, with our current slide deck we could easily fill 3-4 days.
  • Running 5 power strips to tables in the Moscone center costs $2,100. Most of that was $157/hr for the box to plug the power strips into. The room only cost $6K for the day. Methinks I have never been so violated in my life.

The class went well and we learned a heck of a lot. We still have a ton of work to tune the content and package it, but it was awesome to spend a full day teaching folks and getting feedback, as opposed to the usual analyst stuff. I’m starting to think this “cloud” thing might be big.

Today we ran the e10+ program for people with 10+ years in security. I thought we’d delve deeply into technical issues, but they were mostly interested in how to work within their own organizations and prioritize security. To be honest I’m far more comfortable with the pure tech side of things (despite being an analyst), but I do understand that once you hit a certain point in your career the soft skills are more important.

One of my favorite bits from the panel was from Richard Bejtlich. He said one of the ways they determine their priorities is to figure out what the bad guys are looking for. I think I’m going to call this “Attacker Driven Data Classification”. It makes a lot of sense: if the bad guys are looking for something, and it isn’t a high priority for you, at minimum you should figure out why they want it.

Other than that things are going well. We started showing off the Securosis Nexus, which we will make public fairly soon.

With that, it’s time to go to bed. I have two sessions tomorrow (my big DLP presentation and another on cloud and government), plus way too many meetings. Prepping for RSA is always hard, and I hate being away from my family, but it is kind of nice to catch up with folks and be social once (or twice) a year.


Next up, Adrian

What’s new is new.

Rich and Mike put together this year’s e10+ seminar at RSA. And like most panels that involve Securosis, there were a couple testy moments when some of the participants took exception to Richard Bejtlich’s assertion that compromised data is exposed to the world in greater quantity – with far more public access to the content – than ever before. Some of the audience members felt we were seeing the same attacks over and over, and the threats of today are no different than we saw in 1985. In fact they went so far as to say “the cloud” was not much more than publicly available mainframes. David Mortman wins a prize for his rebuttal: “Yeah, RACF rules!”

All kidding aside, I have been in the industry almost that long, and I can say that in some ways this later assertion is true; we still suck at application security, and DoS, non-repudiation, and spoofing work pretty much the same way they did in 1985. How these attacks occur is new, as they exploit both new and old technologies in interesting ways. But the thrust of Richard’s comment is absolutely correct: The speed and quantity of exfiltration is unprecedented. Further, what’s very new is the ability to widely distribute data and make stolen data available for search and inspection. Ten years ago I could push stolen information to FTP servers and hacker sites, but it data was not really accessible to people who did not know where to look for it, or did not understand how to grep through blobs of multi-format data. Now we have Google to do it for us. So what’s old is new again, but in many cases it’s just freakin’ new.


And now for something completely different: Mike

1) We need toddlers, not Kenyans. No offense to Kenyans, but one of the things that became crystal clear at the e10+ sessions this morning was the disparity of needs between the early adopters of security technologies and everyone else. You see the vendors and talking heads spend all their time talking about strategies to do advanced security against serious adversaries. But most of the world can’t even do the simple n00b blocking and tackling to defeat script kiddies.

So basically most of the RSA Conference will be focused on these advanced strategies, which have very little bearing on the vast majority of organizations. It’s like our industry is hiring Kenyans to run a marathon, while everyone else barely walks. Maybe we need schoolteachers.

This capabilities gap might be the most significant issue we face as an industry. Yes, even more than APT or WikiLeaks. Now do two shots, because I said both the buzzword bingo keywords.

2) The perimeter is dead. Long live the perimeter: Besides the fireworks of Cisco and Palo Alto screaming at each other during my panel at the America’s Growth Capital conference, we covered some important ground. First off, the perimeter is not dead. But it needs to get a lot smarter and much more distributed. As more stuff moves to the cloud and video clogs our networks, we need to gain visibility and control, working the areas we know we can access. That means applications.

The other resonating point was the reality that with the massive bandwidth consumed by video, we need to provide network egress as close to the user as we can. That means we need to protect a lot more Internet access points, with more firewall-type devices, and that will strain our management capabilities.

But you already knew that, right?


That brings us to the end of Day 1 at RSA. Will Rich finally get some sleep? Will any of our intrepid heroes spell Richard Bejtlich’s name correctly? Will Mike kidnap some munchkins to help secure the network? Tune in tomorrow to find out.