Category Archives: Hacking

*nix, Hacking, Internet, Internet crime, Security

Blocking IP ranges at the router

Leave a comment

I’m sure that Russia is a wonderful place, and I’m sure that the vast majority of people there are lovely, and have no interest in harming anyone.

Sadly, from the perspective of a server operator, it sometimes feels that nothing good ever comes from Russia.

Being the diligent server operator that I (hopefully) am, I monitor things pretty closely. That includes network traffic coming from the Internet. Over the years, I’ve noticed that a huge proportion of the probes, DDoS attacks, spam, phishing, and hack attempts against my network come from IP addresses in Russia.

It’s gotten to the point where I am now actively blocking huge swathes of Internet addresses (IPs) that originate in Russia and neighbouring countries like the Ukraine.

Blocking those nasty IPs

I run a Linux web server, as well as several Internet-enabled services, at my home office. All of the communications between my server and the Internet pass through a router, making it the ideal place to block unwanted traffic for my entire network, which includes media computers, development systems, and the Windows computer on which I’m writing this.

I’m using a commercial router, but I’ve replaced the original firmware with Advanced Tomato. Doing this provides many benefits, including making it easier to manage the router’s firewall, IPTABLES. Here’s a typical IPTABLES command to block an IP address from the router’s Linux command line:
iptables -I FORWARD -s 185.219.52.90 -j DROP

The DROP directive tells the router to unceremoniously drop any traffic from the specified IP, without logging this action. Traffic can also be logged when it’s dropped, but excessive logging can cause performance problems and fill up logs with junk, so I just drop this traffic.

I issue commands like the one above at my router’s command line to block the traffic immediately, and then I update the router’s startup firewall script with the same command, so that it persists after the next router restart.

So there’s this one guy

There’s been one particularly persistent attacker in the last year or so. This person wants desperately to gain access to one of my Internet-accessible services, but he’s not particularly intelligent, because he keeps trying the same things over and over, in rapid succession. So much so, that at times the traffic he generates comes within shouting distance of a DDoS attack.

I started paying particular attention to traffic associated with a series of ports that are used by the service, and blocking the IP addresses at the other end of that traffic. Whereupon we embarked upon a long game of whac-a-mole, in which I blocked an IP or IP range, and the attacker moved to another host or VPN provider and resumed his attacks from there. It seems clear that this was all being done by one attacker, based on his quick reactions to my blocking.

This went on for several months, but now he appears to have given up. Or at least he’s moved on to other methods.

In the process of blocking all these IPs and networks, the attacker has also helpfully provided me with a list of VPN providers that should be blocked by, well, everyone. Everyone who doesn’t specifically need to allow them.

IP addresses and ranges I’m blocking

Almost all of these IPs and IP ranges are in Russia and the Ukraine. A few are elsewhere in Asia. Most of the ranges are VPN providers.
103.48.51.116
104.129.18.0/23
104.237.192.0/19
104.237.203.0/24
141.98.10.0/24
173.244.208.60
176.67.85.0/24
185.156.72.0/24
185.156.74.0/24
185.193.88.0/24
185.217.69.157
185.219.52.112
185.219.52.90
185.219.52.91
193.106.191.25
193.106.191.35
193.106.191.41
193.32.164.85
193.93.62.0/24
195.54.160.27
198.8.81.220
216.131.114.0/24
216.131.116.0/23
216.131.68.0/24
216.131.88.0/23
217.138.255.202
31.43.185.29
31.43.185.9
37.120.218.0/24
45.134.26.0/24
45.143.203.121
45.145.64.0/23
45.145.65.11
45.146.164.0/23
45.146.166.0/23
45.155.204.0/24
45.155.205.0/24
45.227.253.0/24
45.9.20.0/24
5.188.206.230
71.19.251.0/24
76.180.16.74
77.243.191.120
77.83.36.0/24
78.128.112.18
82.145.32.0/19
84.17.41.141
84.17.41.151
87.251.75.0/24
89.187.182.87
89.187.183.76
91.191.209.110
92.204.240.75
92.255.85.0/24
94.232.40.0/21
98.175.213.148

Here are a few other ranges I’m blocking for various reasons:

  • Hungarian ISP MAGYAR-TELEKOM-MAIN-AS IP range (unceasing garbage): 94.27.128.0/17
  • MediaLand BPH IP range (generally just horrible): 45.141.84.0/24
  • EE-GIGAHOSTINGSERVICES (constant email relay attempts): 176.111.173.0/24
Hacking, Internet crime, Privacy, Security, Things that are bad

Canada Revenue Agency hacked; shuts down online services

Leave a comment

Canadians: if you’ve tried to access your CRA accounts recently, you probably noticed that you can no longer log in. That’s because normal access has been disabled while the CRA works to undo the damage caused by two recent attacks on their services.

The CRA systems were penetrated by persons unknown over the past two weeks. According to the CRA, the breaches have been contained, but the My Account, My Business Account and Represent a Client services have been disabled as a precaution.

Several thousand user accounts have been compromised. Starting in early August, unusual and unauthorized access to accounts was noticed by the account holders and reported to the CRA. In some cases, email, banking, and other account details were changed by the attackers. Fraudulent CERB payments were also issued.

Access to the compromised accounts was apparently gained via ‘credential stuffing’, which is based on the sadly-still-true fact that many people continue to use specific passwords on multiple systems. To be clear: if nobody ever did that, this type of attack would never be successful.

“Of the roughly 12 million active GCKey accounts in Canada, the passwords and usernames of 9,041 users were acquired fraudulently and used to try and access government services, a third of which accessed such services and are being further examined for suspicious activity,” according to a statement from the CRA.

The CRA is in the process of alerting people whose accounts were compromised.

Hacking, Internet crime, Malware, Microsoft, Security, Things that are bad, Windows

Timeline: NSA hacking tool to WannaCry

1 Comment

A recent Washington Post article is helping to answer some questions about Microsoft’s actions in recent months. Here’s a timeline of events:

2012 (or possibly earlier): The NSA identifies a vulnerability in Windows that affects all existing versions of the operating system, and has the potential to allow almost unfettered access to affected systems. A software tool — an exploit — is developed either for, or by, the NSA. The tool is called EternalBlue. People at the NSA worry about the potential damage if the tool or the vulnerability became public knowledge. They decide not to tell anyone, not even Windows’ developer, Microsoft.

EternalBlue finds its way into the toolkit of an elite hacking outfit known as Equation Group. Although it’s difficult to know for certain, this group is generally assumed to be operating under the auspices of the NSA. Equation Group may work for the NSA as contractors, or they may simply be NSA employees. Regardless, the group’s actions seem to align with those of the NSA: their targets are generally in places like Iran, Russia, Pakistan, Afghanistan, India, Syria, and Mali.

Early to mid-2016: A hacking group calling themselves The Shadow Brokers somehow gains access to NSA systems or data, and obtains copies of various NSA documents and tools. Among those tools is EternalBlue.

August, 2016: The Shadow Brokers begin publishing their NSA haul on public services like Tumblr.

January 7, 2017: The Shadow Brokers begin selling tools that are related to EternalBlue.

Late January to early February 2017: The NSA finally tells Microsoft about the vulnerability exploited by EternalBlue. We don’t know exactly when this happened, but it clearly happened. The NSA was Microsoft’s source for this vulnerability.

February 14, 2017: Microsoft announces that February’s Patch Tuesday updates will be postponed. Their explanation is vague: “we discovered a last minute issue that could impact some customers.”

Late February 2017: The Windows SMB vulnerability exploited by EternalBlue is identified publicly as CVE-2017-0144.

March 14, 2017: March’s Patch Tuesday updates from Microsoft include a fix for CVE-2017-0144, MS17-010. The update is flagged as Critical and described as Security Update for Microsoft Windows SMB Server (4013389). Nothing in Microsoft’s output on March 14 calls special attention to this update.

April 14, 2017: The Shadow Brokers release 300 megabytes of NSA material on Github, including EternalBlue.

May 12, 2017: WannaCry ransomware infection wave begins. The malware uses EternalBlue to infect vulnerable computers, mostly Windows 7 PCs in Europe and Asia. Infected computers clearly had not been updated since before March 14, and were therefore vulnerable to EternalBlue.

It’s now clear that the NSA is the real problem here. They had several opportunities to do the right thing, and failed every time, until it was too late. The NSA’s last chance to look at all good in this matter was after the vulnerability was made public, when they should have made the danger clear to the public, or at least to Microsoft. Because, after all, they knew exactly how useful EternalBlue would be in the hands of… just about anyone with bad intent.

Everyone involved in this mess acted foolishly. But whereas we’ve grown accustomed to corporations caring less about people than about money, government institutions — no matter how necessarily secretive — should not be allowed to get away with what the NSA has done. Especially when you consider that this is just the tip of the iceberg. For every WannaCry, there are probably a thousand other threats lurking out there, all thanks to the clowns at the NSA.

Ars Technica’s analysis.

Techdirt’s analysis.

Hacking

Mr. Robot’s realistic depiction of hacking

Leave a comment

As I read Cory Doctorow’s recent Technology Review post, “Mr. Robot Killed the Hollywood Hacker“, I found myself nodding my head enthusiastically. Anyone who knows much about computers and watches Mr. Robot will have noticed that the show’s depiction of hacking is very different from what we usually see on TV and in movies. The user interface is a text console. Everything is done with arcane text commands. Nothing is flashing, except the prompt. In other words, it’s accurate.

For as long as computers have been shown in movies and TV, they have been depicted as flashy, noisy, exploding, and otherwise utterly fanciful, almost magical devices. Hollywood obviously took one look at reality and collectively said “no way, that’s boring as hell.” So the vast majority of computer depictions in movies and TV are some art director’s crazy fantasy of how a computer should look.

I long ago stopped complaining about this. Nobody wants to listen to me drone on about how unrealistic a computer is in some TV show. Now, I just allow myself to be amused. I told myself that this was just harmless hyperbole, a layer of pizazz added onto reality to make it more entertaining.

But Doctorow makes an interesting point: the traditional depiction of computers by Hollywood isn’t as harmless as it may seem.

The 1983 film WarGames is about a high school kid who accidentally hacks into a military computer and almost starts a global thermonuclear war. There were – and still are – a lot of reasons why this is an unlikely scenario, but hey, this is entertainment. A lot of people saw the film, and most enjoyed it, including me. One of the least realistic parts of the film shows the WOPR computer smoking and burning when it becomes overloaded. Real computers just stop working when they’re overtaxed. Anyway, I didn’t mind the silliness; it’s all good fun, right?

Wrong. Apparently, Wargames got some people in Washington worried about whether a high school kid really could hack into military systems and start a war. In 1984, one year after the release of WarGames, the US Congress passed the Computer Fraud and Abuse Act (CFAA), which made activities related to hacking illegal. The legislation is ill-defined, and overly-broad, and it’s widely seen as pointlessly destructive. It contributed to the suicide of Aaron Swartz, who was being charged with crimes related to the CFAA.

The next time you’re watching a TV show or movie, and see an inaccurate depiction of something, ask yourself: “is this really just a harmless Hollywood convenience?”