Security be damned, all in the name of progress.
Today I was able to find a number of IBM i partitions that were live on the Internet with open ports for FTP and Telnet.
I must be one heck of a detective to dig up these open doors, right? How could I tell they were IBM i partitions? How did I find them?
I've been thinking hard about Internet of Things security for the last few months. What really piqued my interest was the incident of police cameras infected with the circa-2008 Conficker worm. Given that many police departments are being outfitted with body cameras and these cameras are used as evidence in criminal prosecution or to absolve police officers from alleged misconduct, the validity of the data and security of the devices must be of paramount concern both to law enforcement and to the public. A body cam with questionable security could be very well thrown out of court as inadmissible evidence, or at least alter the opinions of jurors.
As the world gets more connected, with a predicted 25 billion devices by 2020, one has to wonder how much emphasis is being put on device security, considering the cost of these devices is dropping dramatically. Right now, you can probably buy a webcam for $10 and a wireless home router for $15. Do you really think device security is baked-in with costs that low? What about ongoing support to maintain device security? No way. I've had home routers from major, reputable vendors that haven't seen a firmware update in years because the update is in the new routers, not back-ported to older devices. For me, and only because I try to keep an eye on device security, the old routers get replaced with new ones when the old ones become security-obsolete.
The ability to buy a refrigerator with a built-in webcam and SMTP server to let you know if you need to pick up anything at the grocery store is coming or is already here. Car-based services are incredibly cool, allowing you to know if your mother's tire pressure is running low. We already have smart services in our clothing and our Fitbits and our mobile phones. One has to think, will there be pictures of my kids opening the fridge on the Internet? Will someone be able to know the route patterns my mother takes to drive home? You may remember the 2015 Jeep Cherokee hack that allowed remote operation of the vehicle, which resulted in a 1.4 million unit recall. Carjacking over cellular. Whodathunkit? And the security patch needed to be implemented at a dealer or via a USB stick, so your guess is good as mine just how many of those cars are still out there unprotected. Wireless updates are so easy, yet still it doesn't get done unless it's user-initiated. Imagine having to do a manual USB-based upgrade? Companies are pushing tech at a pace that may not allow the necessary steps to ensure it is secure.
There's a number of IoT search engines that crawl through the Internet, taking snapshots of what they find. One of these search engines is called Shodan. If you get a free account, you can browse anything it digs up for you. One of those easy-to-find things are webcams that operate on port 554. Of course, there are harmless pictures of traffic and weather snapshots from unprotected webcams. Then there are pictures of infants asleep in their cribs.
Even worse. There have been instances in the news of baby monitors being hacked, where parents of toddlers have woken up to strange voices yelling obscenities on their baby monitors.
That's creepy. Especially for us parents out there.
Shodan isn't limited to pictures of webcams. It will take snapshots of whatever open doors it finds. With that being said, let's get back to IBM i.
Using this tool, I was able to easily dig up the IP addresses of some IBM i partitions that were listening on port 21 (FTP). I checked three addresses in total. There's enough evidence in these few examples to prove a point.
Since they had port 21 open, which also gives me the FTP banner, including the DNS name, it's an easy check to see if they had Telnet open. I could also check for HTTP, LDAP, DRDA, or any number of ports just to see what's available to sniff. For the interest of time, I'll stick with what I can find out with ports 21 and 23 open.
Three out of three had plain-text FTP open either through firewall rules or via partitions connected directly to the public Internet. Two out of three had Telnet open. If you've got Telnet open and therefore present a sign-on screen, people now know your system name, which is often the serial number, so they have that also. The serial number comes in real handy in case someone wants to steal a software support call or 20 if they let their own maintenance expire. They also know the city and country. Reverse DNS can give them the domain name, so chances are they know the company name and contact information from WhoIs records.
The amount of information that can be easily deduced from an open FTP port is mind boggling.
Now, if we look at the HelpSystems State of IBM i Security study, which says 33 percent of systems have more than 100 user profiles with default passwords, you might be starting to get the gist of this article. Only 6 percent of companies use the new password rules system value, which allows you to extend your password capabilities. So even if we've changed from the default, 94 percent of us have antiquated rules for those passwords. To be very clear, 94 percent are under the umbrella of V5R4 password rules. Check out this great post from IBM's Dawn May regarding password strength. There's a great section within that highlights the differences between the old and "new" capabilities of IBM i passwords:
The following outlines the IBM i password policy capabilities for version 5 release 4:
- Minimum password length
- Maximum password length
- Restricting new passwords that are the same as old passwords
- Restricting use of consecutive digits
- Requiring a numeric character
- Limiting consecutive repeated alphanumeric characters
- Requiring different characters in the same position of a new password
The additional IBM i password policy capabilities for version 6.1 are:
- Specifying minimum and maximum number of special characters, alphabetic characters, or numeric characters
- Requiring mixed case
- Requiring or disallowing the first character or last character of a password from being a special character or numeric character or alphabetic character
- Disallowing profile name within the password
- Requiring three out of four password components: uppercase characters, lowercase characters, special characters, and numeric characters
Regarding my Shodan search, imagine if one of those partitions had default passwords for user profiles with special authorities. And even if the passwords aren't the default, it's only a matter of time before you could crack it with a brute force attack and have command-line access to FTP. If you've got FTP with a working password, you've got everything else. Even if that password won't allow you Telnet access, you've still got command line through FTP.
Just like the case of web cams that allow an anonymous person to view inside your home, part of the insecurity is weak password strength or just plain default passwords. This isn't really hacking. This is taking advantage of software vulnerabilities and user ignorance. Many people will put a cheap wireless router in their house and never change the default administrator password. If you've left that password as the default, then chances are everything else connected to that router is using the default. If someone can tell the default passwords of a Linksys router by simply using Google, then they can find the default passwords of standard IBM i operating system accounts.
We can only control what we control, so we need to be taking these vulnerabilities seriously. If we put an IBM i service, or any service, on the Internet, it must be secured because if it's on the Internet, it will be found.
The Internet of Things is moving ahead, security be damned. Security starts in our areas of responsibility: our homes and our businesses.