Remember in 2003 when netgear hardcoded a university's NTP server into a ton of their products? Well....<p><a href="https://www.google.com/search?&q=university+ntp+server+netgear" rel="nofollow">https://www.google.com/search?&q=university+ntp+server+netge...</a><p><a href="https://pages.cs.wisc.edu/~plonka/netgear-sntp/" rel="nofollow">https://pages.cs.wisc.edu/~plonka/netgear-sntp/</a>
Apparently they didn't just do that, they also had them make requests way more frequently than they needed to (I guess particularly surprising given that they'd already have to be fairly negligent to have the server hard-coded):<p>> We learned that these packets appeared to be legitimate, well-formed Simple Network Time Protocol (SNTP) version 1 queries, albeit at an inexplicably high rate from each client host. For instance, during one trace, many clients produced about one query per second. This would be highly unusual for a properly constructed SNTP client, since an application which uses SNTP is merely interested in setting its own clock relatively accurately so that its host has some reasonable notion of the current time. One query per second is ridiculous, and is far from best practice for NTP client behavior.
This was a great read, thank you for posting!<p>I love articles like that.
I’m pretty sure this way they’ve hardcoded the NTPs is actually against the ToS for use of the NTP pool too.<p>The way a vendor embedding NTP is _meant_ to do so is documented here:
<a href="https://www.ntppool.org/en/vendors.html" rel="nofollow">https://www.ntppool.org/en/vendors.html</a><p>On another note, back when I ran a web hosting business we hosted a few NTP servers in the pool. It’s such a simple thing to give back, and worth anyone who can make a stable contribution doing so.
> <i>The pool is currently keeping the time of an estimated 5-15 million systems accurate.</i><p>When was this page last updated? I would expect that number to be in the hundreds of millions these days, perhaps even billions.
If you do run an NTP server, please make sure it's not vulnerable to DDoS amplification (monlist, readvar, etc need to be disabled) and apply some rate limiting to make it less useful for reflection attacks. And be proactive about monitoring its traffic volume.<p>If you see high packet rate from a specific IP address or prefix, it's very likely not them abusing your service, but rather you attacking them by responding to spoofed requests.
> The way a vendor embedding NTP is _meant_ to do so is documented here: <a href="https://www.ntppool.org/en/vendors.html" rel="nofollow">https://www.ntppool.org/en/vendors.html</a><p>Note that in the past I've had companies writing embedded linux based firmware using ntppool for time sync request their own vendor zones, however a lot of those requests were ignored so it's unclear if that's still expected. In the end they ended up just using the default ntppool domains since they never got their own vendor zones.
Yes, they should absolutely be using a vendor zone instead of a CNAME under tesla.com.
CNAME'ing pool-ntp.tesla.com to something they do not control is already quite risky as it would allow someone to e.g. request pool-ntp.tesla.com certificate though it might take quite a few tries.
I thought about trying this, but MPIC makes it very very very difficult (the round-robin has some geolocation magic baked in regarding what server it connects you to).
Wouldn't the same apply to pool.ntp.org then?<p>Maybe running a web server on the same IP as an NTP server is a bad idea.
“You must absolutely not use the default pool.ntp.org zone names as the default configuration in your application or appliance.”<p>ref: <a href="https://www.ntppool.org/en/vendors.html" rel="nofollow">https://www.ntppool.org/en/vendors.html</a>
I'd try to contact Assetnote. Most (sadly not all) managed vuln scan companies are pretty sensitive to scanning stuff that doesn't belong to their client and could expose them to liabilities because they don't have permission.
Do a Tesla move and say that's a new service and they need to pay 10000 a month to keep using the service otherwise you reserve the right to mess up with the responses. :P of course, just a joke, but that's what they do
Thankfully it doesn't seem to be much traffic, but still... weird. You'd hope at somepoint the weird responses would get looked at in some log, but I won't hold my breath for that haha.<p>Tangential, but I love the design of your blog. That's so freakishly accurate to old GNOME 2 Ubuntu, amazing work.
So many of the news on HN, such as present one, can be actual stories/scenes from a cyberpunk game/movie these days, that we can safely assume this (otherwise imaginary) future has already arrived.
I really like this website's theme. I noticed this yesterday when using a very old iOS version: Shiny UI is morale-boosting.
As a bug bounty researcher, my systems would do the same thing if they ended up georouted to this IP. *.tesla.com is marked as in scope on <a href="https://bugcrowd.com/engagements/tesla" rel="nofollow">https://bugcrowd.com/engagements/tesla</a>, and my agents will probe anything under there as it is presumed to have explicit authorization.<p>Not sure if there is a great solution, but I'm inclined to say that attack traffic like this is the new normal. In fact, the attack volume they got is quite small compared to the volume I have seen on other tech company subdomains - the new normal is probably much worse.
To OP robinpie:<p>I understand that Tesla is treating your NTP server, a volunteer server and part of the greater volunteer pool of NTP servers, as their own infrastructure.<p>However, I can't tell from the article if the scans originate with:<p>A.) IT staff at Tesla that are scanning exposed services on what they perceive, or claim wrongly, as their own network for vulnerabilities.<p>B.) Somehow a rogue operator (read botnet)<p>C.) A rogue operator who is using the cars themselves to run exploit scans?<p>C would be the most alarming and concerning.
I just checked my own webserver logs (I also run a sever in the ntp pool) and I too see some hits in my webserver logs.<p>They look to all be log4j vuln scanning activity (CVE-2021-44228), and the volume isn't that high (a few a day, and not every day). They just have some overzealous vuln scanning. And yes, they shouldn't have the NTP pool under their DNS name.<p>I've had all sorts of strange things happen because of my ntp pool membership, this one is pretty far on the benign end of things.
> rg -zFI pool-ntp.tesla.com access.log*<p>Neat! Didn’t know about this command that’s very helpful
I've been consistently attacked by ShadowServer who have the following sponsors,<p>Akamai,
APNIC Foundation,
Arctic Security,
AusCERT,
Avast,
Backblaze,
Canadian Center for Cyber Security,
CERT.AT,
CERT.br,
CERT.LV,
CIRA,
CIRCL,
Craig Newmark Philanthropies,
CSIRT.LI,
CSIS Security Group,
DFN‑CSIRT,
Digital Trust Center,
EURid,
HelseCERT,
ICANN,
Identity Digital,
KPN,
Mastercard,
NASK (CERT.pl),
NCSC Ireland,
NICS,
Nihon Cyber Defence,
Nucleus Security,
Orange Polska,
Precursor Security,
Protect.ngo,
Public Interest Registry (PIR),
Red Hat,
SURFcert,
SWITCH,
Team Cymru,
Trend Micro,
Trivest AG,
Tucows,
Verisign,
VulnCheck,<p>I don't care what they say they're doing, I hate how corporations can act with impunity with these types of things while everyone else would get a felony for it.
Calling vuln scanning from a non-profit a felony is a bit of a stretch. Many for-profit companies do similar vuln scanning and then threaten companies with security "scorecards". That is borderline extortion.
If the non-profit was walking down the road and rattling everybody’s door lock to see which are unlocked, and having a look around the windows to see if any are open, would that be a crime?<p>Because that is exactly what all of these vulnerability scanning companies are doing, and all of us sort of just… let them.
If the neighborhood was constantly being canvased by criminals checking doorknobs, so your concerned neighbor went over to your house to check your doorknob, and then let you know if you accidentally left it open, would you also accuse your neighbor of being a criminal trying to break in?
Yes.<p>I have personally done this before, the correct sequence of moves is:<p>1. Call your neighbor, ask for permission.
2. Check the door
3. Lock the door<p>If you don’t have their phone number, you are not on good enough terms to touch their lock.
This is why the checking doors / neighborhood analogy isn't a good one.<p>Having one person with poor computer security negatively impacts everyone. Hacked sites turn into phishing landing pages, exploit kit hosting, stolen data dumps, and launching off points for attacks on everyone else. The vuln scanning ShadowServer is doing is meant to be a public good, which is why they share the info with ISPs and governments.<p>Security is too intertwined to stand by and say other people's vulnerabilities aren't your problem.
It is a good analogy because thieves stealing from one house successfully gives them resources and incentivizes them stealing from the same area again.<p>And even if you remove the analogy, ShadowServer means good, but good intentions doesn’t necessarily make their action moral or legal.<p>Yes, compromised servers can be used by hackers as means to commit crimes. But when these groups scan the entire internet, they do cause harm as well, as shown in the original linked article. Much less harm than a black hat, but they still waste time and resources from innocent third parties.<p>It’s fair to ask if the harm they cause is worth the good they do.
Report it <a href="https://www.cisa.gov/reporting-cyber-incident" rel="nofollow">https://www.cisa.gov/reporting-cyber-incident</a>
Can't you report them to the FBI, someone is clearly trying to hack your server
I would assume the fastest way to actually make this stop would be to setup a bunch of honeypot exploits, trigger their detection and someone will figure out what they did wrong.<p>Other than not, with these huge companies you have 0 recourse.
I did something like this a few weeks ago on my photography site: <a href="https://robertmay.photography/journal/meta-has-tried-to-scrape-this-site-1-million-times-in-2-weeks-ive-given-them-toasters-instead" rel="nofollow">https://robertmay.photography/journal/meta-has-tried-to-scra...</a><p>Meta not only hasn't noticed, but is currently sending about 11 requests per second to my site. I've also seemingly trapped one of those TV proxy scraper nets as I'm getting absolutely hammered by requests from all over the place now. I get maybe 10 legit visitors per day, and I'm currently blocking 406,787 IPs from things that have fallen into my honeypot.<p>I've tweaked my site to return empty status responses a configurable amount of time but the traffic has been so intense that Traefik is now struggling, so I'm going to have to figure out something else. I was returning over-capacity errors and I think that was a mistake, I've swapped to 400 range status codes now. I don't want to use Cloudflare so I'm not sure what to do after this.<p>The people at these companies are either incompetent or malicious.
Return a HTTP 301 pointing to <a href="https://facebook.com" rel="nofollow">https://facebook.com</a>?<p>Might make them scan themselves instead.
Since it's a photography site, route 'em over to goatse. That might get someone's attention.
I actually think this is brilliant. Let the scanner IP get access to the most unpatched Wordpress, maybe even generate some garbage ERP numbers.<p>Once the asset scanner detects the vulns, everyone will kick into high gear to patch this.
I wonder if source IP can be ping triangulated, then total ping count displayed as "Tesla MAU: +/-x% today", "Suspected ownership changes this month: xxx cars" by apparent home location changes, then residential and Tesla-unrelated locations excluded server side, and plotted on the map, then finally the whole system exposed to the public Internet and shared to SpaceX fans.<p>"Hey Texas Model S #345 just left KXYZ, moving at >100mph towards the pad. Everyone get cameras out!"<p>It'll be gone by lunchtime that day.
(OOP here). I wonder if switching my replies from 299 to 200 OK would be enough for some of them.
Unlikely, but the lazy wonks could enable stratum 1 time services with the cars GPS/cellular receiver and RTC. Having done that for equipment in the past, we all know it is literary only $8.43/unit in parts. Also, setting ntp time as the tertiary fall back has been around for over a decade. =3
Often, the folks you want to ban are not the hosts running the scans.<p>One's best bet is to play possum, and use your clients last login IPs falling in your service area geo-IP ranges for a firewall white-list. Then redirect the other traffic for a black hole route.<p>If the nuisance hosts assume they have driven the host offline, they will eventually give up and move on. =3
Tangential; I love the style drop down.
<i>> unintentionally be being a nuisance.</i><p>As opposed to <i>intentionally</i> being a nuisance?
It’s just laziness on Tesla’s part because working your bare bones crew to failure isn’t the exception but the SOP. Same why you can’t reach anyone about Supercharger issues, this is by design to run as lean as possible for profits.<p>I see this in cyber clients that are perpetually short staffed and simply accept the degraded system because the consequences are minimal. If nothing important breaks, nothing changes.
I bet we'll find that its openAI agent benchmarks off on their own again
I'd report this to AWS instead since Tesla doesn't seem aware. It's AWS IPs as the source.
I might try Searchlight Cyber first (it's their service doing this, presumably without their knowledge), and then if that doesn't work go to AWS.
Yeah it looks like someone thinks they're "hacking" Tesla when really it's hitting volunteer IPs.
Give Tesla the wrong time.
Huh, are we sure this is just NTP shenanigans and not some super duper dangerous AI agents "escaping" the "containment"?
Is this basically a free pen test, or am I misunderstanding the setup?
50,000 requests in the span of ... 21 days? It might be interesting if it were 50 million in those 21 days (even then, 27 RPS is ... nothing). Drop a report to AWS T&S, block, drop, reject, and move on. This is like less than background noise at this point.<p>Edit: tamping down a bit of my prickliness because it looks like this individual is a relative newcomer to running internet-facing services. This is actually a pretty good intro to that: the place as a whole is a cesspool and any conceivable “attack”, scan, probe, pentest is, has, or will be happening at all times. Some you can mitigate yourself, others you’ll need to bring in a specialist service (see DDoS sinks and mitigation services, for example) or contact someone’s abuse address, others you just have to ignore because it’s just not directed at you personally, or just not worth more than three seconds’ thought beyond a firewall rule. This is the latter. Maybe interesting if this is the first time you’re seeing something like this, but for more grizzled operators like myself, it doesn’t even register as notable anymore.
[dead]
[dead]
> They tried all kinds of exploits against me ... probing WordPress and other CMS management endpoints<p>This is standard bot crawler traffic. Anyone who runs a home server sees attempts to load wp paths all the time
In case of the very likely title only reader posting:<p>> Speculation: Assetnote pulled in everything it could find under tesla.com, including pool-ntp.tesla.com, which CNAMEs to pool.ntp.org, which can resolve to my machine — 67.215.249.229. The asset inventory saves this as a Tesla asset, and starts throwing exploits at me, a stranger.<p>> Not a vuln in Tesla, and I'm not asking for anything, but I just wanted to let you know that you may unintentionally be being a nuisance.
This seems like a gross exaggeration.<p>This happens EVERY day to EVERY web server out there. I have a personal site that gets thousands of requests per day from bots.<p>Running a public server (like NTP) means you will get tons of strange requests. Moreso if you run a web server on the same IP because bots will scrape certificate transparency logs. The entire IPv4 space is scanned continuously.<p>This may sound harsh, but you cannot stop it. It is whack-a-mole. Filter it and move on, go outside and touch grass, seriously. This is not worth being upset over.<p>I treat these as an opportunity to tune my filters and firewall rules.
You don’t think there’s a difference between “hackers try to attack everything“ and “Tesla decided that I personally need to be tested as one of their systems due to a lazy misconfiguration“ are different?
Are you asking if I think Elon had this person's website on an 80" screen in a conference room whilst they collude how to send traffic to their site?<p>No, I do not.<p>I get thousands of these "security scanner" requests on a low-traffic site weekly, if not daily. Not a day goes by where someone doesn't try to kick in my door. Are you prepared to waste valuable time in your life to separate the white hat from the black hat requests?<p>I would have added the IP addresses to my firewall's IP blacklist and forgotten about it the next day. It's really all you can do.
I'm aware how much crap there is on the Internet, I just think the specific nature of this (legitimate commercial vuln scanner thinks I'm Tesla) is funny
> it has received ~8,000 requests from two of your scanning hosts<p>If it were 8000 requests <i>per second</i>, this might be worthy of some investigation.<p>But 8000 ntp requests alone consume far less than 1 us cent of compute + bandwidth. This isn't worth lifting a finger over.
Please read the article, it's not the volume of the NTP requests, they're actively sending exploit/attempt to compromise payloads. They're probing things in a way that you would ordinarily only do to your own internal infrastructure.<p>"They tried all kinds of exploits against me: path traversal, webshell uploads, probing software internals, probing WordPress and other CMS management endpoints, SSRF, Log4Shell, and a lot more."
Isn't this technically a crime, since they're actively attempting to access a computer system they don't own?
yup just report them to the FBI
In today’s world, a crime is only a crime if you get charged. Tesla has enough power to not get charged.
Tesla isn’t doing the scanning though, instead somebody thinks they are scanning Tesla, but Tesla points them to someone else.<p>The scanner is likely illegal.<p>The pointing is… so stupid nobody thought to make a law about it.
It's HTTP requests, not NTP requests, and the volume isn't the problem, it's that Assetnote is sending live exploit payloads /at all/ to a stranger on Tesla's behalf
It's not 8000 requests. It's 8000 attempts to exploit various software on OP's server.