Last Friday, my workday was rudely interrupted because I couldn’t access Github. To add insult to injury I couldn’t even complain about it on Twitter. I tried to drown my sorrows by listening to moody Leonard Cohen songs on Spotify, but alas…
You’ve probably heard of this. Huge tracts of the Internet were down because the DNS provider Dyn faced a massive Denial of Service attack from the Mirai botnet, which takes advantage of Internet of Things devices like cameras and DVRs.
“Oh & also, evil sorcerers crippled our divination network Friday by getting millions of coffee makers & lightswitches to shout real loud” https://t.co/OPhdOJLwc1— Max Gladstone (@maxgladstone) October 25, 2016
So, what’s new about Mirai?
Can you guess how Mirai spreads?
Was there new zeroday in the devices? Hey, maybe there was an old, unpatched vulnerability hanging — who has time to apply software updates to their toaster? Maybe it was HeartBleed 👻?
Mirai does one, and only one thing in order to break into new devices: it cycles through a bunch of default username/password combinations over telnet, like “admin/admin” and “root/realtek”. For a laugh, “mother/fucker” is in there too.
Default credentials. Over telnet. That’s how you get hundreds of thousands of devices. The Morris worm from 1988 tried a dictionary password attack too, but only after its buffer overflow and sendmail backdoor exploits failed.
Oh, and Morris’ password dictionary was larger, too.
How do we keep getting this wrong?
Around the world, we spend $75 billion a year on information security. And for what, when we keep getting such basic things wrong? Suppose I waved a magic wand and cut the worldwide security budget in half. Would things really be that much worse? The security industry is addicted to selling expensive complicated products instead of doing the basics well.
I was at a security conference the other week, and there was yet another crop of cyberapocalypse talks. The Internet of Things is a garbage fire. Industrial control systems are going to get us all killed. Users are clicking phishing links like sheep. We’re all doomed. And somehow, it’s always the fault of shitty programmers or dumb users. Let’s all laugh at their fails.
It’s all bullshit.
We sell biometric authentication systems to people who need a good password manager. We sell live threat attribution intelligence with colorful maps to people who need to practice configuration management. We sell advanced in-cpu sandbox endpoint protection to people who need to institute a patching program. There’s a reason why security practitioners get such a kick out of ThreatButt.
There are lots of real, important, conceptually difficult problems in security. We don’t really know how to write secure code, and it’s all too easy to get socially engineered. But, right now, the vast majority of threats can be thwarted by the basics:
- Keep your systems patched
- Keep your systems properly configured.
- Make sure you have strong passwords and two factor authentication.
Do the basics first. The basics matter. Then you can focus on the Sisyphean tasks that remain. Instead, here we are selling fancy bullshit and barely making any progress in 28 years. Lots of money in it, though.
Paying the Bills
Surprise, I also sell a security product! But I will say this: Appcanary isn’t going to protect you from shipping millions of internet-accessible cameras with the same password. We won’t even protect you from having your DNS provider DoSed.
The major botnet of 2016 is simpler than the botnet of 1988. There’s something wrong in how we do security, and at Appcanary, we think it’s a complete lack of focus on the basics.
The highest value, easiest thing you can do to improve your security is patch known vulnerabilities. Most breaches come from years-old vulnerabilities.
Our product, Appcanary, monitors your apps and servers, and notifies you whenever a new vulnerability is discovered in a package you rely on.
Sign up today!