Tankless systems have the same corrosion problem. They also have the added benefit of scale reducing their effectiveness and eventually clogging them entirely. If your tanked heater was properly installed -- it has a catch pan and drain -- then the eventual pinhole leaks will not be the ruination you claim. Either way, preventive maintenance is necessary to keep them running for years. (my previous one lasted 23 years! The flue eventually burned through -- short of spraying it with the stuff used to protect welding tips, not much I can do to protect the steel from the fire that's supposed to heat it.)
(the ones at my parent's house (electric) have always rusted from the inside out because no one replaces the cathode rods)
who the fuck has a water heater that is network connectable
Mine can be -- iCOMM industrial control system. It's in my house, so it doesn't have one. And it wouldn't be "connected to the internet" if I did, but it would be connected to my leads-to-the-internet LAN.
Having seen numerous documentaries and tours of the UK National Grid, and various US grid operators, nobody runs at exactly 100%. A healthy grid is run with some excess capacity as a buffer -- while small, for obvious cost reductions, it's important to have that buffer to avoid brownouts due to sudden increases is load. (i.e. the exact thing the hacker is trying to cause.) I'll pick on the UK because their load has the grandest predictable spikes... when many "click on the kettle".
Because most people dumped Solaris more than a decade ago?
"Not entirely" as in [in my experience] many former solaris shops still have bits of solaris remaining. It wasn't "dumped", but incrementally replaced over the years. (solaris 10 was the real kick-in-the-ass to start moving... SMF, the systemD of the Solaris world.)
OpenSolaris was as much a marketing ploy as it was a means to remain relevant -- "Open Source" being the trendy new buzz word / business model. You could already get solaris for free -- for "non-commercial" use. (surprisingly, even under the infinite greed of Oracle, solaris is still available for free.) Sun was the only source for sparc hardware to run the OS, so they already had your money. (solaris/x86 never had much of a software market)
(Sure, anyone running x86 hardware is far better off moving to linux or even windows. The few solaris/x86 installs I knew were moved to linux and windows.)
Not entirely. I still run across the odd solaris 7/8/9 system from time to time. I still run one myself. You don't fuck with what isn't broken; there hasn't been a need to replace it, although it has been discussed. (maybe should based entirely on the power bill)
(I have several systems standing by for testing and troubleshooting whatever might come across my desk. But they aren't actually on.)
Incorrect. A growing amount of Cisco hardware is running linux. Old School IOS and Old School PIX aren't linux -- which shouldn't be a surprise as they pre-date linux, but modern ASA, NX-OS, IOS-XR/XE have a linux base. (XR started out with QNX and moved to linux ~5yr ago)
Yes, software switches do exist (aka "bridge"), but, as you mention, they're slow as crap because software (general purpose CPU) has to move frames from interface to interface.
Amazon isn't "getting into the hardware biz". They're just going to (sub)contract that shit to any number of "white box" switch makers already gluing common Broadcom (etc.) switch SoCs to boards. The OS on those boxes will most likely just be a customized / rebadged existing network OS.
SDN is still mostly just Marketing Lies(tm). The only people to really do it, have been doing it much longer than the term has existed. And they do it with in-house designed technology that Works For Them(tm) -- and they generally don't share. (facebook and rackspace claim to opensource their shit. Good luck trying to use what little they've shared.)
Cisco bought Linksys to get it's name into the consumer market. It failed. All it did was tarnish the name "Cisco" in the enterprise market, and significantly confuse people who don't know the difference.
At each dispenser. They look like a tiny odometer at the top of the terminal. Sometimes you can hear them click. They are not connected to any electronic thing that I'm aware of. (eg. they have to be manually recorded)
As almost every station has buried tanks these days, I don't know if they have a meter on the flow pump itself.
I always laugh at that one, too. I'll give the most relatable car analogy... "how long do the wipers on your car last?" If that doesn't get a nod, I offer to dump fine grained sand on their windshield and let them see how that goes.:-)
If your blog actually is private, then go right ahead. You're free to have whatever opinions and views you like. It's only when you start sharing them in public that it becomes other people's business.
Indeed. I'm sorry Ms. Price completely missed the point. This isn't a matter of "if reddit wants you fired". If one's actions in public reflect poorly on their employer, don't be surprised when they fire you.
Crash the control systems with the valves open, then there's nothing to (a) meter what's being pumped, (b) report anything being pumped, or (c) close that valve. You'd have to walk out there with a key, open the unit, and reset it. (or flip a breaker to power off all or part of the station) And that's possibly not a key the clerk has. (there are two doors: one for the receipt printer, and one for the computer)
I can't say about MI, but in NC, every pump I've ever seen has a mechanical flow counter... for tax purposes. I can't say the station is recording those numbers on any scale that would help here. (daily, hourly...) If you crash the control interface, the back-office systems won't have a count, but the tax man will!
The "at least 10" part most likely comes from an estimate of security footage.
100L ~ 25gal, so you don't have anything magical. Yes, most autos have a ~20gal tank. This ain't Mad Max; we don't need 100gal tanks. That said, many trucks (pickups) and SUVs will have something on the upper end of the scale, but they also have pretty poor fuel economy.
We recently had a car in a neighbor's lot catch fire. (they didn't lock it, and some heroin idiots set it on fire) A fire truck was there in ~2min, they were actively dealing with it in under 5min. That car was right behind a gas station, so they were likely highly motivated to stop that explosion. Yes, it did take ~20min for a cop to show up to block off the street.
There have been times when I've called the police and it took hours for someone to finally respond. Non-emergencies are zero priority. (they had wrecks, burglaries, and drunks to deal with)
A malfunctioning gas pump that's giving away free gas isn't remotely an emergency. There are at least a dozen different ways to deal with the situation. From turning off the entire station (there's actually only one pump (per grade); the things from which you get the gas is a just a metered dispenser), to padlocking the pump, to parking other cars around that pump to keep others away from it.
There are plenty of ways to secure BGP, and routing in general. However, just like the locks on your house, they don't do you any good if you don't actually lock them. We have yet to see a BGP session be hijacked, or an external attacker inject a rogue route into an established BGP session. What we DO see all the time are flaming idiots accepting whatever the hell someone advertises.
Done. And Done. They took over the address space for Amazon's DNS service (Route 53), so they ARE the DNS for many domains. That gives them 100% control of all DNS answers, including where the server is. That traffic now goes to a server they control. It's trivial to get a Let's Encrypt signed certificate under these conditions.
(Of course, these guys didn't even bother to do that.)
Nope. The issue (ssl certificate) is still entirely a Big Giant Fail(tm) on Let's Encrypt's part. If I can take over your DNS, I can effectively become your server and *poof* now I can those fools to sign a certificate for my stolen domain. Now, these guys didn't actually do that, so there actions where immediately evident.
Yes, they used BGP to announce more specific routes to parts of Amazon's DNS infrastructure so that traffic came to them. They were then in effective control of many domains, but apparently chose to hijack some cryptocurrency site.
There are plenty of ways to secure BGP, and routing in general. However, just like the locks on your house, they don't do you any good if you don't actually lock them. We have yet to see a BGP session be hijacked, or an external attack inject a rogue route into an established BGP session. What we DO see all the time are flaming idiots accepting whatever the hell someone advertises.
Simple. Those aren't new subway lines. They're extensions to, and interconnects between existing subway lines. There's been a lot of construction over the decades, but in total, the entire system is still quite old.
Slashdot Mirror
Tankless systems have the same corrosion problem. They also have the added benefit of scale reducing their effectiveness and eventually clogging them entirely. If your tanked heater was properly installed -- it has a catch pan and drain -- then the eventual pinhole leaks will not be the ruination you claim. Either way, preventive maintenance is necessary to keep them running for years. (my previous one lasted 23 years! The flue eventually burned through -- short of spraying it with the stuff used to protect welding tips, not much I can do to protect the steel from the fire that's supposed to heat it.)
(the ones at my parent's house (electric) have always rusted from the inside out because no one replaces the cathode rods)
Mine can be -- iCOMM industrial control system. It's in my house, so it doesn't have one. And it wouldn't be "connected to the internet" if I did, but it would be connected to my leads-to-the-internet LAN.
Having seen numerous documentaries and tours of the UK National Grid, and various US grid operators, nobody runs at exactly 100%. A healthy grid is run with some excess capacity as a buffer -- while small, for obvious cost reductions, it's important to have that buffer to avoid brownouts due to sudden increases is load. (i.e. the exact thing the hacker is trying to cause.) I'll pick on the UK because their load has the grandest predictable spikes... when many "click on the kettle".
"Not entirely" as in [in my experience] many former solaris shops still have bits of solaris remaining. It wasn't "dumped", but incrementally replaced over the years. (solaris 10 was the real kick-in-the-ass to start moving... SMF, the systemD of the Solaris world.)
OpenSolaris was as much a marketing ploy as it was a means to remain relevant -- "Open Source" being the trendy new buzz word / business model. You could already get solaris for free -- for "non-commercial" use. (surprisingly, even under the infinite greed of Oracle, solaris is still available for free.) Sun was the only source for sparc hardware to run the OS, so they already had your money. (solaris/x86 never had much of a software market)
(Sure, anyone running x86 hardware is far better off moving to linux or even windows. The few solaris/x86 installs I knew were moved to linux and windows.)
Not entirely. I still run across the odd solaris 7/8/9 system from time to time. I still run one myself. You don't fuck with what isn't broken; there hasn't been a need to replace it, although it has been discussed. (maybe should based entirely on the power bill)
(I have several systems standing by for testing and troubleshooting whatever might come across my desk. But they aren't actually on.)
Incorrect. A growing amount of Cisco hardware is running linux. Old School IOS and Old School PIX aren't linux -- which shouldn't be a surprise as they pre-date linux, but modern ASA, NX-OS, IOS-XR/XE have a linux base. (XR started out with QNX and moved to linux ~5yr ago)
running on top of linux... on top of commodity merchant (*cough*broadcom*cough*) silicon.
Yes, software switches do exist (aka "bridge"), but, as you mention, they're slow as crap because software (general purpose CPU) has to move frames from interface to interface.
Amazon isn't "getting into the hardware biz". They're just going to (sub)contract that shit to any number of "white box" switch makers already gluing common Broadcom (etc.) switch SoCs to boards. The OS on those boxes will most likely just be a customized / rebadged existing network OS.
SDN is still mostly just Marketing Lies(tm). The only people to really do it, have been doing it much longer than the term has existed. And they do it with in-house designed technology that Works For Them(tm) -- and they generally don't share. (facebook and rackspace claim to opensource their shit. Good luck trying to use what little they've shared.)
Cisco bought Linksys to get it's name into the consumer market. It failed. All it did was tarnish the name "Cisco" in the enterprise market, and significantly confuse people who don't know the difference.
At each dispenser. They look like a tiny odometer at the top of the terminal. Sometimes you can hear them click. They are not connected to any electronic thing that I'm aware of. (eg. they have to be manually recorded)
As almost every station has buried tanks these days, I don't know if they have a meter on the flow pump itself.
I always laugh at that one, too. I'll give the most relatable car analogy... "how long do the wipers on your car last?" If that doesn't get a nod, I offer to dump fine grained sand on their windshield and let them see how that goes. :-)
If your blog actually is private, then go right ahead. You're free to have whatever opinions and views you like. It's only when you start sharing them in public that it becomes other people's business.
Indeed. I'm sorry Ms. Price completely missed the point. This isn't a matter of "if reddit wants you fired". If one's actions in public reflect poorly on their employer, don't be surprised when they fire you.
#1 doesn't work anymore. The pump will end the transaction after 30-60s with no flow.
Crash the control systems with the valves open, then there's nothing to (a) meter what's being pumped, (b) report anything being pumped, or (c) close that valve. You'd have to walk out there with a key, open the unit, and reset it. (or flip a breaker to power off all or part of the station) And that's possibly not a key the clerk has. (there are two doors: one for the receipt printer, and one for the computer)
I can't say about MI, but in NC, every pump I've ever seen has a mechanical flow counter... for tax purposes. I can't say the station is recording those numbers on any scale that would help here. (daily, hourly...) If you crash the control interface, the back-office systems won't have a count, but the tax man will!
The "at least 10" part most likely comes from an estimate of security footage.
100L ~ 25gal, so you don't have anything magical. Yes, most autos have a ~20gal tank. This ain't Mad Max; we don't need 100gal tanks. That said, many trucks (pickups) and SUVs will have something on the upper end of the scale, but they also have pretty poor fuel economy.
We recently had a car in a neighbor's lot catch fire. (they didn't lock it, and some heroin idiots set it on fire) A fire truck was there in ~2min, they were actively dealing with it in under 5min. That car was right behind a gas station, so they were likely highly motivated to stop that explosion. Yes, it did take ~20min for a cop to show up to block off the street.
There have been times when I've called the police and it took hours for someone to finally respond. Non-emergencies are zero priority. (they had wrecks, burglaries, and drunks to deal with)
A malfunctioning gas pump that's giving away free gas isn't remotely an emergency. There are at least a dozen different ways to deal with the situation. From turning off the entire station (there's actually only one pump (per grade); the things from which you get the gas is a just a metered dispenser), to padlocking the pump, to parking other cars around that pump to keep others away from it.
In a drill, I can understand that. Go where you're supposed to go. In an actual emergency, do whatever is necessary.
Play God?!? One person is not "god". One review will not get you fired. A PATTERN of bad reviews will get you fired.
There are plenty of ways to secure BGP, and routing in general. However, just like the locks on your house, they don't do you any good if you don't actually lock them. We have yet to see a BGP session be hijacked, or an external attacker inject a rogue route into an established BGP session. What we DO see all the time are flaming idiots accepting whatever the hell someone advertises.
Done. And Done. They took over the address space for Amazon's DNS service (Route 53), so they ARE the DNS for many domains. That gives them 100% control of all DNS answers, including where the server is. That traffic now goes to a server they control. It's trivial to get a Let's Encrypt signed certificate under these conditions.
(Of course, these guys didn't even bother to do that.)
Nope. The issue (ssl certificate) is still entirely a Big Giant Fail(tm) on Let's Encrypt's part. If I can take over your DNS, I can effectively become your server and *poof* now I can those fools to sign a certificate for my stolen domain. Now, these guys didn't actually do that, so there actions where immediately evident.
Yes, they used BGP to announce more specific routes to parts of Amazon's DNS infrastructure so that traffic came to them. They were then in effective control of many domains, but apparently chose to hijack some cryptocurrency site.
There are plenty of ways to secure BGP, and routing in general. However, just like the locks on your house, they don't do you any good if you don't actually lock them. We have yet to see a BGP session be hijacked, or an external attack inject a rogue route into an established BGP session. What we DO see all the time are flaming idiots accepting whatever the hell someone advertises.
Simple. Those aren't new subway lines. They're extensions to, and interconnects between existing subway lines. There's been a lot of construction over the decades, but in total, the entire system is still quite old.