Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Water system controllers don't belong on the internet, says ex-NSA chief

theregister.com

Post reply on HN
Print view View original

151–160 of 163 posts

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#151

Post by black6 »

giantg2 wrote: post #149
spauldo wrote: post #114

Earlier quoted context omitted.

A microcontroller based solution is rarely a good replacement for a PLC. You can find integrators willing to come out and troubleshoot your device in the field a lot more easily than embedded progammers and vendor support is usually excellent. When your plant is down, that sort of thing matters. PLCs handle harsh environments for decades, whereas with a microcontroller all the thermal issues and power conditioning ar…

"but you won't need to do it again for another quarter century." Hmm, so how do they handle the new cyber threats, since most of these are connected? It seems upgrades would need to be performed to address new threats. And just because something is not internet connected doesn't make it immune (eg stuxnet).

They shouldn't be connected is the point of the article. All PLC and DCS manufacturers are constantly working on threat mitigation and pushing out updates to address vulnerabilities. It's on the Process Controls team to decide what is an actual threat that requires patching.

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#153

Post by palmotea »

thyristan wrote: post #141
palmotea wrote: post #120

Earlier quoted context omitted.

> You mean like, just hang a thumb drive on a lanyard inside of the cabinet -- with source code and such? This guy is very happy when someone literally does that: https://www.youtube.com/@Cursed_Controls (great channel btw), because usually he has to deal with worse. > And if it includes the programming software, then that lets Joe (from over in shipping) have a go at rejiggering the packaging machine. That's a probl…

The PLC is usually in a control cabinet that can be opened by a control cabinet key: https://www.knipex.de/schaltschrankschluessel?v=2921 Real locks in such systems are rare, sometimes even disallowed. Usually, a lock with a real key is only used to temporarily prevent switching on a machine during maintenance or such.

> The PLC is usually in a control cabinet that can be opened by a control cabinet key:

As far as I can tell, that thing's $50. Why would Joe from shipping have one?

> Usually, a lock with a real key is only used to temporarily prevent switching on a machine during maintenance or such.

Most locks aren't really meant to resist and adversary. They're meant to communicate who is allowed and who is not, and back that up by making violations a little non-trivial.

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#154

Post by analog31 »

hiAndrewQuinn wrote: post #136
analog31 wrote: post #69

Earlier quoted context omitted.

>>> The pay is also so much worse in my experience. The result is that anybody who can program well enough to develop software is doing so. Those with any programming skills who remain behind are smart-but-undisciplined programmers, people who have a reason to be in a particular geographic area, too old to think about changing careers, or just plain nuts. This is what "lack of investment in manufacturing, research, a…

To me that speaks to a fundamentally incorrect economic structure. I bet PLC programmers could command much higher salaries - at much higher levels of risk - if they held more liability for things going wrong. That, in turn, would spur a lot of economic development around unfucking the field. On the other hand, markets are generally smarter than I am. Much like how US citizens living in the EU are basically barred fr…

Deming wrote, "quality control is a management problem." When you see a plant where workers are cutting corners on quality, they're probably under pressure to do so.

Factories face liability already, either due to the cost of "line down" or getting sued. They could pass this liability on to their employees, who would probably cover it through insurance, like doctors.

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#155

Post by thyristan »

palmotea wrote: post #153
thyristan wrote: post #141

Earlier quoted context omitted.

The PLC is usually in a control cabinet that can be opened by a control cabinet key: https://www.knipex.de/schaltschrankschluessel?v=2921 Real locks in such systems are rare, sometimes even disallowed. Usually, a lock with a real key is only used to temporarily prevent switching on a machine during maintenance or such.

> The PLC is usually in a control cabinet that can be opened by a control cabinet key: As far as I can tell, that thing's $50. Why would Joe from shipping have one? > Usually, a lock with a real key is only used to temporarily prevent switching on a machine during maintenance or such. Most locks aren't really meant to resist and adversary. They're meant to communicate who is allowed and who is not, and back that up b…

That's the first search engine hit I've pasted. The cheap ones that work just as well are single-digit prices. And usually there is one in every toolbox on the shop floor.

Edit: if you need proof, 2 for 6.50€, shipping included: https://www.ebay.de/itm/277671545941

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#156

Post by amluto »

spauldo wrote: post #114
amluto wrote: post #61

Earlier quoted context omitted.

Another odd observation about the industry. For better or for worse, I inherited a little system controlled by a Fanuc PLC. It had a bug and would occasionally get stuck in a bad state. By some minor miracle, I managed to track down the person who had originally programmed it (now retired and moved out of state), and he emailed me a file and introduced me to his apprentice (who lives in a third state). I found a tria…

A microcontroller based solution is rarely a good replacement for a PLC. You can find integrators willing to come out and troubleshoot your device in the field a lot more easily than embedded progammers and vendor support is usually excellent. When your plant is down, that sort of thing matters. PLCs handle harsh environments for decades, whereas with a microcontroller all the thermal issues and power conditioning ar…

This system barely deserves a PLC. The modules are only a couple of extremely basic analog inputs and outputs, plus a wildly outdated and unsupported HMI. And the entire plant could possibly be replaced with something more modern for less money than upgrading the PLC. (Or I could get something cheap from AutomationDirect - I have no idea how good their PLCs are.)

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#157

Post by bigfatkitten »

1970-01-01 wrote: post #3

He is wrong and right. They should be connected to the Internet when they aren't 30 year old PLCs ripe for abuse. Until then, cut the data lines and do water monitoring the old way.

Things that don’t have a compelling need to be connected to the internet should never be connected to the internet. When there is a need, connectivity should be restricted to the bare minimum required for the task.

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#158

Post by bigfatkitten »

gertrunde wrote: post #115
mikewarot wrote: post #68

Earlier quoted context omitted.

A data diode can allow monitoring via the internet without risking ingress of control. Commercial units aren't cheap though. Obviously we need open source designs. Perhaps the easiest way would be a Raspberry pi set up with an opto isolated CGA/EGA/VGA/SVGA capture that could be viewed via the internet? (I mean, we're probably talking systems still running MS-DOS or Windows 98 running these systems)

If you want to do that sort of thing in a cheap way, then you can. I'm told a box with two optical NICs in it with only one fibre in it, TX->RX, works quite well, plus a bit of software to wrap around it. But these sorts of organisations love a bit of bureaucratic regulation, and "certified" devices and that sort of thing, so there's probably not much of a market for a data diode without the paperwork. On the opto is…

Data diode hardware isn’t terribly expensive, but the software runs into the hundreds of thousands precisely because these are niche products.

To DIY it you need three NICs (or an optical isolator, which is honestly more work than it’s worth) because the transmitting NIC needs to see a link up before it transmits anything. That link is most easily provided by the second port on the transmitting side.

Developing software to reliably move data across the link, and ensure that it’s properly formatted etc will still be your problem.

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#159

Post by AngryData »

lokar wrote: post #24
grebc wrote: post #4

Earlier quoted context omitted.

If it’s connected, it’s compromised. Or will be. Folly to think otherwise.

A lot of water infrastructure is physically spread out. It be very expensive and cumbersome (and probably inefficient) to require staff to by physically present at each site for monitoring and making any changes.

Its not like we lack the labor to do it. Plus how many years or decades of wages for someone to stand around would be required to exceed the cost of infastructure going down even just a few days?

Re: Water system controllers don't belong on the internet, says ex-NSA chief

#160

Post by thrill »

closeparen wrote: post #103
judge2020 wrote: post #71

Earlier quoted context omitted.

Alternatively, small drones exist, and so do low power devices you could slingshot/shoot into a secure area and set to auto wipe after they've done their job. Maybe could even build it out of biodegradable material so it'll clean up in short time.

If you've got secret agents on enemy soil who can get near targets with drones, you're probably just bombing them.

Why bomb something, which carries its own risk profile and success criteria, when you can make some piece of equipment malfunction just enough to wear out in a year instead of 20 years?
Post reply on HN