Slashdot Mirror
Hospital Equipment Infected With Conficker
nandemoari writes "Recently, the Conficker/Downadup worm infected several hundred machines and critical medical equipment in an undisclosed number of US hospitals.
The attacks were not widespread; however, Marcus Sachs, director of the SANS Internet Storm Center, told CNET News that it raises the awareness of what we would do if there were millions of computers infected in hospitals or in critical infrastructure locations.
It's not clear how the devices (including heart monitors, MRI machines and PCs) got infected. Infected computers were running Windows NT and Windows 2000 in a local area network (LAN) that wasn't supposed to be Internet accessible, but the LAN was connected to one with direct Internet access.
A patch was released by Microsoft last October that fixes the problem, but the computers infected were reportedly too old to be patched."
Newer isn't always better.
Why risk having security vulnerabilities on a tried and tested mission-critical system? They should have gone with Linux or BSD from the start and had virtually guaranteed upgrade compatibility from that point on, with plenty of commercial support options.
Sam ty sig.
1) Vendors of these devices almost across the board disallow local IT admins to put any windows patches on the machines
- this is due to FDA requirements for approval, and the vendor is "covering" themselves
- also, they usually have a list of "qualified updates" that is usually MONTHS behind MS's patch cycle (not surprising given the sheer number and speed of holes that are found)
- usually the vendors claim that THEY will apply patches regularly, in practice, they almost NEVER do
2) Vendors typically disallow these machines to be on the active directory
- this is because they can't stand troubleshooting/supporting issues in their software due to GPO's being pushed down, software management software, etc etc
3) To everyone screaming how idiotic it is that medical devices have Windows on them: you may be a geek, but have clearly never worked in a real enterprise environment. Windows is embedded on so many devices in the world (medical and otherwise) that you would never even know existed. Why? Because it's widely supported, has huge hardware support, and is surprisingly OPEN to developers to hack it into whatever they need it to be. And windows programmers are a dime a dozen.
4) To everyone screaming how idiotic it is that medical devices are connected to the internet getting infected - Do you even know how Conficker spreads? It spreads quite easily across a LAN, attaching to Windows file shares. See MS08-067 for more info. Many of these devices are on a LAN with no DNS (although plenty are on the 'net). Why? Again, because vendors insist that they be connected so they can VPN in and support them (often using LogMeIn, Webex etc).
THE MAGIC WORDS ARE SQUEAMISH OSSIFRAGE
but the LAN was connected to one with direct Internet access.
Internet enabled machine got infected, and bridged over to the closed-off network. Why SMB was enabled on the embedded systems is a better question.
The article says "A patch was released by Microsoft last October ..." The availability of a patch doesn't mean squat. Before a patch can bve installed on medical equipment, the hardware vendor has to validate the patch. In other words, the vendor has to test the ever loving crap out of the software to insure it does not conpromise patient safety.
The fact that cornflicker got on life safety and mission critical systems at all raises the question of why anyone would use a consumer grade operating system for mission critical systems or for life support systems. At a minimum, these systems should have been running Unix or Solaris. Vx Works or Linux are also good, but require a higher level of computer engineed to implement.
This is just plain lunacy.
Does it bother anyone else that "critical medical equipment" was running Windows NT or 2000?
Of course it does. Building any medical equipment around an intrinsically unreliable system is about as irresponsible a decision as anyone could make.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Medical equipment has a very long lifespan. Many devices for measurement and monitoring are used for 10 to 20 years before replacement. The general policy is "if it works, don't fix it and, more important, do not touch it". The real problem is that most suppliers of equipment are reluctant to support any type of patches. Many of the suppliers explicitly state that the machines may not be changed in any way (and that includes patching the OS) or you will lose all guarantee and support.
Doesn't Microsoft itself say (perhaps in the EULA disclaimer) that its operating systems were not intended to be used in this sort of mission-critical capacity? That could of course have a very narrow definition, something along the lines of "don't ever use it to operate that iron lung but maybe use it so the receptionist can run MS Office" but if that were the case, then this would be a mere nuisance and not such a real problem. That is, in that case there'd be nothing special about the fact that the affected institution happened to be a hospital beyond the fact that it sounds bad. Because of that, I really get the impression that they were using the wrong tool for the job.
It is a miracle that curiosity survives formal education. - Einstein
The biggest issue here is that Medical Equipment has to be run through an FDA Validation process. If you make changes to the system, you have to revalidate, and Validation takes months and $100K's. So the vendors leave them as-is.
What's frustrating is that these systems need to be on a LAN, since they need to report their results to other clinical systems. So these small islands need to be linked other islands, and eventually, someone screws up and links an island with an Internet connection . . . .
I don't mean to nitpick, but what's the difference? Your ISP has a network that's hooked to the Internet and you connect your computer to it in order to have Internet access. Seems to me that the basic routing functionality of IP guarantees that there is no meaningful difference there, at least not unless you have some carefully-planned firewall rules in place and even then ...
It is a miracle that curiosity survives formal education. - Einstein
Oh, really?
When operating in direct electron-beam therapy mode, a low-powered electron beam was emitted directly from the machine, then spread to safe concentration using scanning magnets. When operating in megavolt X-ray mode, the machine was designed to rotate four components into the path of the electron beam: a target, which converted the electron beam into X-rays; a flattening filter, which spread the beam out over a larger area; a set of movable blocks (also called a collimator), which shaped the X-ray beam; and an X-ray ion chamber, which measured the strength of the beam.
The accidents occurred when the high-power electron beam was activated instead of the intended low power beam, and without the beam spreader plate rotated into place. The machine's software did not detect that this had occurred, and therefore did not prevent the patient from receiving a potentially lethal dose of radiation. The high-powered electron beam struck the patients with approximately 100 times the intended dose of radiation, causing a feeling described by patient Ray Cox as "an intense electric shock". It caused him to scream and run out of the treatment room.[2] Several days later, radiation burns appeared and the patients showed the symptoms of radiation poisoning. In three cases, the injured patients died later from radiation poisoning.
The software flaw is recognized as a race condition.
It's extremely cynical of me perhaps, but I wonder if this isn't some type of pernicious planned obsolesence. Some car makers for many years deliberately made cars to last 20,000 hours (pure folklore, overheard) because they needed cars to fail after a few years to keep the volume of new car sales going.
Wouldn't the same principle work with computers? Something has to make them fail over time or people will make do with the old. Unfortunate that this means NT4 boxes in hospitals might get people killed, but when have the truly greedy ever really cared?
Do not mock my vision of impractical footwear
The real problem is that most suppliers of equipment are reluctant to support any type of patches. Many of the suppliers explicitly state that the machines may not be changed in any way (and that includes patching the OS) or you will lose all guarantee and support.
Shouldn't they be using OpenBSD, then?
Why are you getting modded as "Funny?" That's the first thought I had. Shouldn't heart monitors and MRI machines have an embedded OS of some sort? MRIs are more complex, but (AFAIK) things like heart monitors do one thing and one thing only.
This signature serves no purpose other than to help you see which posts were made by me.
Because the network admin should have the laboratory equipment firewalled off with a "deny all" preceded and followed by comment lines that read " # DANGER -- MEDICAL EQUIPMENT ON THIS SEGMENT -- If you permit so much as one stinking port to pass through this firewall, I will hunt you down and leave nothing behind for the doctors to patch together."
There is no excuse on the planet for letting health care equipment see the cloud. If data has to enter or leave, it should pass through a bastion host. If the requirements are that the equipment really has to reach the internet, the requirements are faulty.
John
Yes, an air gap would be even better, but I read the above to say "we had an air gap but some idiot bridged it." My intent was to provide the bridge as part of the design to let people know that you've already created the path, but to then plaster it with the equivalent of "Road Closed", "Keep Out", and "No Trespassing" signs.
John