document/TranslateProject
2
0
Fork 0
mirror of https://github.com/LCTT/TranslateProject.git synced 2025-01-25 23:11:02 +08:00
Code Issues Packages Projects Releases Wiki Activity

Translated: Internet Chemotherapy

Browse Source
This commit is contained in:
Yixun Xu 2018-01-30 21:51:39 -05:00
parent 5b0079e5ed
commit faf8cc1344
2 changed files with 76 additions and 336 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

336
sources/tech/20171218 Internet Chemotherapy.md
View File

@ -1,336 +0,0 @@
(yixunx translating)
Internet Chemotherapy
======
12/10 2017
### 1. Internet Chemotherapy
Internet Chemotherapy was a 13 month project between Nov 2016 - Dec 2017.
It has been known under names such as 'BrickerBot', 'bad firmware
upgrade', 'ransomware', 'large-scale network failure' and even
'unprecedented terrorist actions.' That last one was a little harsh,
Fernandez, but I guess I can't please everybody.
You can download the module which executes the http and telnet-based
payloads from this router at http://91.215.104.140/mod_plaintext.py. Due to
platform limitations the module is obfuscated single threaded python, but
the payloads are in plain view and should be easy to figure out for any
programmer worth his/her/hir salt. Take a look at the number of payloads,
0-days and techniques and let the reality sink in for a moment. Then
imagine what would've happened to the Internet in 2017 if I had been a
blackhat dedicated to building a massive DDoS cannon for blackmailing the
biggest providers and companies. I could've disrupted them all and caused
extraordinary damage to the Internet in the process.
My ssh crawler is too dangerous to publish. It contains various levels of
automation for the purpose of moving laterally through poorly designed
ISP networks and taking them over through only a single breached router.
My ability to commandeer and secure hundreds of thousands of ISP routers
was the foundation of my anti-IoT botnet project as it gave me great
visibility of what was happening on the Internet and it gave me an
endless supply of nodes for hacking back. I began my non-destructive ISP
network cleanup project in 2015 and by the time Mirai came around I was
in a good position to react. The decision to willfully sabotage other
people's equipment was nonetheless a difficult one to make, but the
colossally dangerous CVE-2016-10372 situation ultimately left me with no
other choice. From that moment on I was all-in.
I am now here to warn you that what I've done was only a temporary band-
aid and it's not going to be enough to save the Internet in the future.
The bad guys are getting more sophisticated, the number of potentially
vulnerable devices keep increasing, and it's only a matter of time before
a large scale Internet-disrupting event will occur. If you are willing to
believe that I've disabled over 10 million vulnerable devices over the 13-
month span of the project then it's not far-fetched to say that such a
destructive event could've already happened in 2017.
YOU SHOULD WAKE UP TO THE FACT THAT THE INTERNET IS ONLY ONE OR TWO
SERIOUS IOT EXPLOITS AWAY FROM BEING SEVERELY DISRUPTED. The damage of
such an event is immeasurable given how digitally connected our societies
have become, yet CERTs, ISPs and governments are not taking the gravity
of the situation seriously enough. ISPs keep deploying devices with
exposed control ports and although these are trivially found using
services like Shodan the national CERTs don't seem to care. A lot of
countries don't even have CERTs. Many of the world's biggest ISPs do not
have any actual security know-how in-house, and are instead relying on
foreign vendors for help in case anything goes wrong. I've watched large
ISPs withering for months under conditioning from my botnet without them
being able to fully mitigate the vulnerabilities (good examples are BSNL,
Telkom ZA, PLDT, from time to time PT Telkom, and pretty much most large
ISPs south of the border). Just look at how slow and ineffective Telkom
ZA was in dealing with its Aztech modem problem and you will begin to
understand the hopelessness of the current situation. In 99% of the
problem cases the solution would have simply been for the ISPs to deploy
sane ACLs and CPE segmentation, yet months later their technical staff
still hasn't figured this out. If ISPs are unable to mitigate weeks and
months of continuous deliberate sabotage of their equipment then what
hope is there that they would notice and fix a Mirai problem on their
networks? Many of the world's biggest ISPs are catastrophically negligent
and this is the biggest danger by a landslide, yet paradoxically it
should also be the easiest problem to fix.
I've done my part to try to buy the Internet some time, but I've gone as
far as I can. Now it's up to you. Even small actions are important. Among
the things you can do are:
* Review your own ISP's security through services such as Shodan and take
them to task over exposed telnet, http, httpd, ssh, tr069 etc. ports on
their networks. Refer them to this document if you have to. There's no
good reason why any of these control ports should ever be accessible
from the outside world. Exposing control ports is an amateur mistake.
If enough customers complain they might actually do something about it!
* Vote with your wallet! Refuse to buy or use 'intelligent' products
unless the manufacturer can prove that the product can and will receive
timely security updates. Find out about the vendor's security track
record before giving them your hard-earned money. Be willing to pay a
little bit more for credible security.
* Lobby your local politicians and government officials for improved
security legislation for IoT (Internet of Things) devices such as
routers, IP cameras and 'intelligent' devices. Private or public
companies currently lack the incentives for solving this problem in the
immediate term. This matter is as important as minimum safety
requirements for cars and general electrical appliances.
* Consider volunteering your time or other resources to underappreciated
whitehat organizations such as GDI Foundation or Shadowserver
Foundation. These organizations and people make a big difference and
they can significantly amplify the impact of your skillset in helping
the Internet.
* Last but not least, consider the long-shot potential of getting IoT
devices designated as an 'attractive nuisance' through precedent-
setting legal action. If a home owner can be held liable for a
burglar/trespasser getting injured then I don't see why a device owner
(or ISP or manufacturer) shouldn't be held liable for the damage that
was caused by their dangerous devices being exploitable through the
Internet. Attribution won't be a problem for Layer 7 attacks. If any
large ISPs with deep pockets aren't willing to fund such precedent
cases (and they might not since they fear that such precedents could
come back to haunt them) we could even crowdfund such initiatives over
here and in the EU. ISPs: consider your volumetric DDoS bandwidth cost
savings in 2017 as my indirect funding of this cause and as evidence
for its potential upside.
### 2. Timeline
Here are some of the more memorable events of the project:
* Deutsche Telekom Mirai disruption in late November 2016. My hastily
assembled initial TR069/64 payload only performed a 'route del default'
but this was enough to get the ISP's attention to the problem and the
resulting headlines alerted other ISPs around the world to the
unfolding disaster.
* Around January 11-12 some Mirai-infected DVRs with exposed control port
6789 ended up getting bricked in Washington DC, and this made numerous
headlines. Gold star to Vemulapalli for determining that Mirai combined
with /dev/urandom had to be 'highly sophisticated ransomware'. Whatever
happened to those 2 unlucky souls in Europe?
* In late January 2017 the first genuine large-scale ISP takedown occured
when Rogers Canada's supplier Hitron carelessly pushed out new firmware
with an unauthenticated root shell listening on port 2323 (presumably
this was a debugging interface that they forgot to disable). This epic
blunder was quickly discovered by Mirai botnets, and the end-result was
a large number of bricked units.
* In February 2017 I noticed the first Mirai evolution of the year, with
both Netcore/Netis and Broadcom CLI-based modems being attacked. The
BCM CLI would turn out to become one of the main Mirai battlegrounds of
2017, with both the blackhats and me chasing the massive long tail of
ISP and model-specific default credentials for the rest of the year.
The 'broadcom' payloads in the above source may look strange but
they're statistically the most likely sequences to disable any of the
endless number of buggy BCM CLI firmwares out there.
* In March 2017 I significantly increased my botnet's node count and
started to add more web payloads in response to the threats from IoT
botnets such as Imeij, Amnesia and Persirai. The large-scale takedown
of these hacked devices created a new set of concerns. For example,
among the leaked credentials of the Avtech and Wificam devices there
were logins which strongly implied airports and other important
facilities, and around April 1 2017 the UK government officials
warned of a 'credible cyber threat' to airports and nuclear
facilities from 'hacktivists.' Oops.
* The more aggressive scanning also didn't escape the attention of
civilian security researchers, and in April 6 2017 security company
Radware published an article about my project. The company trademarked
it under the name 'BrickerBot.' It became clear that if I were to
continue increasing the scale of my IoT counteroffensive I had to come
up with better network mapping/detection methods for honeypots and
other risky targets.
* Around April 11th 2017 something very unusual happened. At first it
started like so many other ISP takedowns, with a semi-local ISP called
Sierra Tel running exposed Zyxel devices with the default telnet login
of supervisor/zyad1234. A Mirai runner discovered the exposed devices
and my botnet followed soon after, and yet another clash in the epic
BCM CLI war of 2017 took place. This battle didn't last long. It
would've been just like any of the hundreds of other ISP takedowns in
2017 were it not for something very unusual occuring right after the
smoke settled. Amazingly, the ISP didn't try to cover up the outage as
some kind of network issue, power spike or a bad firmware upgrade. They
didn't lie to their customers at all. Instead, they promptly published
a press release about their modems having been vulnerable which allowed
their customers to assess their potential risk exposure. What did the
most honest ISP in the world get for its laudable transparency? Sadly
it got little more than criticism and bad press. It's still the most
depressing case of 'why we can't have nice things' to me, and probably
the main reason for why 99% of security mistakes get covered up and the
actual victims get left in the dark. Too often 'responsible disclosure'
simply becomes a euphemism for 'coverup.'
* On April 14 2017 DHS warned of 'BrickerBot Threat to Internet of
Things' and the thought of my own government labeling me as a cyber
threat felt unfair and myopic. Surely the ISPs that run dangerously
insecure network deployments and the IoT manufacturers that peddle
amateurish security implementations should have been fingered as the
actual threat to Americans rather than me? If it hadn't been for me
millions of us would still be doing their banking and other sensitive
transactions over hacked equipment and networks. If anybody from DHS
ever reads this I urge you to reconsider what protecting the homeland
and its citizens actually means.
* In late April 2017 I spent some time on improving my TR069/64 attack
methods, and in early May 2017 a company called Wordfence (now Defiant)
reported a significant decline in a TR069-exploiting botnet that had
previously posed a threat to Wordpress installations. It's noteworthy
that the same botnet temporarily returned a few weeks later using a
different exploit (but this was also eventually mitigated).
* In May 2017 hosting company Akamai reported in its Q1 2017 State of the
Internet report an 89% decrease in large (over 100 Gbps) DDoS attacks
compared with Q1 2016, and a 30% decrease in total DDoS attacks. The
largest attack of Q1 2017 was 120 Gbps vs 517 Gbps in Q4 2016. As large
volumetric DDoS was one of the primary signatures of Mirai this felt
like concrete justification for all the months of hard work in the IoT
trenches.
* During the summer I kept improving my exploit arsenal, and in late July
I performed some test runs against APNIC ISPs. The results were quite
surprising. Among other outcomes a few hundred thousand BSNL and MTNL
modems were disabled and this outage become headline news in India.
Given the elevated geopolitical tensions between India and China at the
time I felt that there was a credible risk of the large takedown being
blamed on China so I made the rare decision to publically take credit
for it. Catalin, I'm very sorry for the abrupt '2 day vacation' that
you had to take after reporting the news.
* Previously having worked on APNIC and AfriNIC, on August 9th 2017 I
also launched a large scale cleanup of LACNIC space which caused
problems for various providers across the subcontinent. The attack made
headlines in Venezuela after a few million cell phone users of Movilnet
lost service. Although I'm personally against government surveillance
of the Internet the case of Venezuela is noteworthy. Many of the
LACNIC ISPs and networks have been languishing for months under
persistent conditioning from my botnet, but Venezuelan providers have
been quick to fortify their networks and secure their infrastructure.
I believe this is due to Venezuela engaging in far more invasive deep
packet inspection than the other LACNIC countries. Food for thought.
* In August 2017 F5 Labs released a report called "The Hunt for IoT: The
Rise of Thingbots" in which the researchers were perplexed over the
recent lull in telnet activity. The researchers speculated that the
lack of activity may be evidence that one or more very large cyber
weapons are being built (which I guess was in fact true). This piece
is to my knowledge the most accurate assessment of the scope of my
project but fascinatingly the researchers were unable to put two and
two together in spite of gathering all the relevant clues on a single
page.
* In August 2017 Akamai's Q2 2017 State of the Internet report announces
the first quarter in 3 years without the provider observing a single
large (over 100 Gbps) attack, and a 28% decrease in total DDoS attacks
vs Q1 2017. This seems like further validation of the cleanup effort.
This phenomenally good news is completely ignored by the mainstream
media which operates under an 'if it bleeds it leads' mentality even
when it comes to information security. This is yet another reason why
we can't have nice things.
* After the publication of CVE-2017-7921 and 7923 in September 2017 I
decided to take a closer look at Hikvision devices, and to my horror
I realized that there's a technique for botting most of the vulnerable
firmwares that the blackhats hadn't discovered yet. As a result I
launched a global cleanup initiative around mid-September. Over a
million DVRs and cameras (mainly Hikvision and Dahua) were disabled
over a span of 3 weeks and publications such as IPVM.com wrote several
articles about the attacks. Dahua and Hikvision wrote press releases
mentioning or alluding to the attacks. A huge number of devices finally
got their firmwares upgraded. Seeing the confusion that the cleanup
effort caused I decided to write a quick summary for the CCTV people at
http://depastedihrn3jtw.onion.link/show.php?md5=62d1d87f67a8bf485d43a05ec32b1e6f
(sorry for the NSFW language of the pastebin service). The staggering
number of vulnerable units that were online months after critical
security patches were available should be the ultimate wakeup call to
everyone about the utter dysfunctionality of the current IoT patching
process.
* Around September 28 2017 Verisign releases a report saying that DDoS
attacks declined 55% in Q2 2017 vs Q1, with a massive 81% attack peak
decline.
* On November 23rd 2017 the CDN provider Cloudflare reports that 'in
recent months, Cloudflare has seen a dramatic reduction in simple
attempts to flood our network with junk traffic.' Cloudflare speculates
it could've partly been due to their change in policies, but the
reductions also line up well with the IoT cleanup activities.
* At the end of November 2017 Akamai's Q3 2017 State of the Internet
report sees a small 8% increase in total DDoS attacks for the quarter.
Although this was a significant reduction compared to Q3 2016 the
slight uptick serves as a reminder of the continued risks and dangers.
* As a further reminder of the dangers a new Mirai strain dubbed 'Satori'
reared its head in November-December of 2017. It's particularly
noteworthy how quickly the botnet managed to grow based on a single
0-day exploit. This event underlines the current perilous operating
state of the Internet, and why we're only one or two severe IoT
exploits away from widespread disruption. What will happen when nobody
is around to disable the next threat? Sinkholing and other whitehat/
'legal' mitigations won't be enough in 2018 just like they weren't
enough in 2016. Perhaps in the future governments will be able to
collaborate on a counterhacking task force with a global mandate for
disabling particularly severe existential threats to the Internet, but
I'm not holding my breath.
* Late in the year there were also some hysterical headlines regarding a
new botnet that was dubbed 'Reaper' and 'IoTroop'. I know some of you
will eventually ridicule those who estimated its size at 1-2 million
but you should understand that security researchers have very limited
knowledge of what's happening on networks and hardware that they don't
control. In practice the researchers could not possibly have known or
even assumed that most of the vulnerable device pool had already been
disabled by the time the botnet emerged. Give the 'Reaper' one or two
new unmitigated 0-days and it'll become as terrifying as our worst
fears.
### 3. Parting Thoughts
I'm sorry to leave you in these circumstances, but the threat to my own
safety is becoming too great to continue. I have made many enemies. If
you want to help look at the list of action items further up. Good luck.
There will also be those who will criticize me and say that I've acted
irresponsibly, but that's completely missing the point. The real point
is that if somebody like me with no previous hacking background was able
to do what I did, then somebody better than me could've done far worse
things to the Internet in 2017. I'm not the problem and I'm not here to
play by anyone's contrived rules. I'm only the messenger. The sooner you
realize this the better.
-Dr Cyborkian a.k.a. janit0r, conditioner of 'terminally ill' devices.
--------------------------------------------------------------------------------
via：https://ghostbin.com/paste/q2vq2
作者：janit0r
译者：[译者ID](https://github.com/译者ID)
校对：[校对者ID](https://github.com/校对者ID)
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，
[Linux中国](https://linux.cn/) 荣誉推出

76
translated/tech/20171218 Internet Chemotherapy.md Normal file
View File

@ -0,0 +1,76 @@
互联网化疗
======
译注：本文作者 janit0r 被认为是 BrickerBot 病毒的作者。此病毒会攻击物联网上安全性不足的设备并使其断开和其他网络设备的连接。janit0r 宣称他使用这个病毒的目的是保护互联网的安全，避免这些设备被入侵者用于入侵网络上的其他设备。janit0r 称此项目为”互联网化疗“。janit0r 决定在 2017 年 12 月终止这个项目，并在网络上发表了这篇文章。
12/10 2017
### 1. 互联网化疗
互联网化疗是在 2016 年 11 月 到 2017 年 12 月之间的一个为期 13 个月的项目。它曾被称为“BrickerBot”、“错误的固件升级”、“勒索软件”、“大规模网络瘫痪”，甚至“前所未有的恐怖行为”。最后一个有点伤人了，费尔南德斯（译注：委内瑞拉电信公司 CANTV 的光纤网络曾在 2017 年 8 月受到病毒攻击，公司董事长曼努埃尔·费尔南德斯称这次攻击为[“前所未有的恐怖行为”][1]），但我想我大概不能让所有人都满意吧。
你可以从 http://91.215.104.140/mod_plaintext.py 下载我的代码模块，它可以基于 http 和 telnet 发送恶意请求（译注：这个链接已经失效，不过在 [Github][2] 上有备份）。因为平台的限制，模块里是混淆过的单线程 Python 代码，但请求的内容依然是明文，任何合格的程序员应该都能看得懂。看看这里面有多少恶意请求、0-day 漏洞和入侵技巧，花点时间让自己接受现实。然后想象一下，如果我是一个黑客，致力于创造出强大的 DDoS 生成器来勒索那些最大的互联网服务提供商和公司的话，互联网在 2017 年会受到怎样的打击。我完全可以让他们全部陷入混乱，并同时对整个互联网造成巨大的伤害。
我的 ssh 爬虫太危险了，不能发布出来。它包含很多层面的自动化，可以只利用一个被入侵的路由器就能够在设计有缺陷的互联网服务提供商的网络上平行移动并加以入侵。正是因为我可以征用数以万计的互联网服务提供商的路由器，而这些路由器让我知晓网络上发生的事情并给我源源不断的节点用来进行入侵行动，我才得以进行我的反物联网僵尸网络项目。我于 2015 年开始了我的非破坏性的提供商网络清理项目，于是当 Mirai 病毒入侵时我已经做好了准备来做出回应。主动破坏其他人的设备仍然是一个困难的决定，但无比危险的 CVE-2016-10372 漏洞让我别无选择。从那时起我就决定一不做二不休。
（译注：上一段中提到的 Mirai 病毒首次出现在 2016 年 8 月。它可以远程入侵运行 Linux 系统的网络设备并利用这些设备构建僵尸网络。本文作者 janit0r 宣称当 Mirai 入侵时他利用自己的 BrickerBot 病毒强制将数以万计的设备从网络上断开，从而减少 Mirai 病毒可以利用的设备数量。）
我在此警告你们，我所做的只是权宜之计，它并不足以在未来继续拯救互联网。坏人们正变得更加聪明，可能有漏洞的设备数量在持续增加，发生大规模的、能使网络瘫痪的事件只是时间问题。如果你愿意相信我曾经在一个持续 13 个月的项目中使上千万有漏洞的设备变得无法使用，那么不过分地说，如此严重的事件本可能在 2017 年就发生。
__你们应该意识到，只需要再有一两个严重的物联网漏洞，我们的网络就会严重瘫痪。__ 考虑到我们的社会现在是多么依赖数字网络，而计算机安全应急响应组和互联网服务提供商们又是多么地忽视问题的严重性，这种事件造成的伤害是无法估计的。互联网服务提供商在持续地部署有开放的控制端口的设备，而且即使像 Shodan 这样的服务可以轻而易举地发现这些问题，我国的计算机安全应急响应组还是似乎并不在意。而很多国家甚至都没有自己的计算机安全应急响应组。世界上许多的互联网服务提供商都没有雇佣任何熟知计算机安全问题的人，而是在出现问题的时候依赖于外来的专家来解决。我曾见识过大型互联网服务提供商在我的僵尸网络的调节之下连续多个月持续受损，但他们还是不能完全解决漏洞（几个好的例子是 BSNL、Telkom ZA、PLDT、某些时候的 PT Telkom，以及南半球大部分的大型互联网服务提供商）。只要看看 Telkom ZA 解决他们的 Aztech 调制解调器问题的速度有多慢，你就会开始理解现状有多么令人绝望。在 99% 的情况下，要解决这个问题只需要互联网服务提供商部署合理的访问控制列表并把部署在用户端的设备单独分段，但是几个月过去之后他们的技术人员依然没有弄明白。如果互联网服务提供商在经历了几周到几个月的针对他们设备的蓄意攻击之后仍然无法解决问题，我们又怎么能期望他们会注意到并解决 Mirai 在他们网络上造成的问题呢？世界上许多最大的互联网服务提供商对这些事情无知得令人发指，而这毫无疑问是最大的危险，但奇怪的是，这应该也是最容易解决的问题。
我已经尽自己的责任试着去让互联网多坚持一段时间，但我已经尽力了。接下来要交给你们了。即使很小的行动也是非常重要的。你们能做的事情有：
* 使用像 Shodan 之类的服务来检查你的互联网服务提供商的安全性，并驱使他们去解决他们网络上开放的 telnet、http、httpd、ssh 和 tr069 等端口。如果需要的话，可以把这篇文章给他们看。从来不存在什么好的理由来让这些端口可以从外界访问。开放控制端口是业余人士的错误。如果有足够的客户抱怨，他们也许真的会采取行动！
* 用你的钱包投票！拒绝购买或使用任何“智能“产品，除非制造商保证这个产品能够而且将会收到及时的安全更新。在把你辛苦赚的钱交给提供商之前，先去查看他们的安全记录。为了更好的安全性，可以多花一些钱。
* 游说你本地的政治家和政府官员，让他们改进法案来规范物联网设备，包括路由器、IP 照相机和各种”智能“设备。不论私有还是公有的公司目前都没有足够的动机去在短期内解决问题。这件事情和汽车或者通用电气的安全标准一样重要。
* 考虑给像 GDI 基金会或者 Shadowserver 基金会这种缺少支持的白帽黑客组织贡献你的时间或者其他资源。这些组织和人能产生巨大的影响，并且他们可以很好地发挥你的能力来帮助互联网。
* 最后，虽然希望不大，但可以考虑通过设立法律先例来让物联网设备成为一种”<ruby>诱惑性危险品<rt>attractive nuisance</rt></ruby>“（译注：attractive nuisance 是美国法律中的一个原则，意思是如果儿童在私人领地上因为某些对儿童有吸引力的危险物品而受伤，领地的主人需要负责，无论受伤的儿童是否是合法进入领地）。如果一个房主可以因为小偷或者侵入者受伤而被追责，我不清楚为什么设备的主人（或者互联网服务提供商和设备制造商）不应该因为他们的危险的设备被远程入侵所造成的伤害而被追责。连带责任原则应该适用于对设备应用层的入侵。如果任何有钱的大型互联网服务提供商不愿意为设立这样的先例而出钱（他们也许的确不会，因为他们害怕这样的先例会反过来让自己吃亏），我们甚至可以在这里还有在欧洲为这个行动而众筹。互联网服务提供商们：把你们在用来应对 DDoS 的带宽上省下的可观的成本当做我为这个目标的间接投资，也当做它的好处的证明吧。
### 2. 时间线
下面是这个项目中一些值得纪念的事件：
* 2016 年 11 月底的德国电信 Mirai 事故。我匆忙写出的最初的 TR069/64 请求只执行了 `route del default`，不过这已经足够引起互联网服务提供商去注意这个问题，而它引发的新闻头条警告了全球的其他互联网服务提供商来注意这个迫近的危机。
* 大约 1 月 11 日 到 12 日，一些位于华盛顿特区的开放了 6789 控制端口的硬盘录像机被 Mirai 入侵并瘫痪，这上了很多头条新闻。我要给 Vemulapalli 点赞，她居然认为 Mirai 加上 /dev/urandom 一定是“非常复杂的勒索软件”（译注：Archana Vemulapalli 当时是华盛顿市政府的 CTO）。欧洲的那两个可怜人又怎么样了呢？
* 2017 年 1 月底发生了第一起真正的大规模互联网服务提供商下架事件。Rogers Canada 的提供商 Hitron 非常粗心地推送了一个在 2323 端口上监听的无验证的 root shell（这可能是一个他们忘记关闭的 debug 接口）。这个惊天的失误很快被 Mirai 的僵尸网络所发现，造成大量设备瘫痪。
* 在 2017 年 2 月，我注意到 Mirai 在这一年里的第一次扩张，Netcore/Netis 以及基于 Broadcom CLI 的调制解调器都遭受了攻击。BCM CLI 后来成为了 Mirai 在 2017 年的主要战场，黑客们和我自己都在这一年的余下时间里花大量时间寻找无数互联网服务提供商和设备制造商设置的默认密码。前面代码中的“broadcom”请求内容也许看上去有点奇怪，但它们是统计角度上最可能禁用那些大量的有问题的 BCM CLI 固件的序列。
* 在 2017 年 3 月，我大幅提升了我的僵尸网络的节点数量并开始加入更多的网络请求。这是为了应对包括 Imeij、Amnesia 和 Persirai 在内的僵尸网络的威胁。大规模地禁用这些被入侵的设备也带来了新的一些问题。比如在 Avtech 和 Wificam 设备所泄露的登录信息当中，有一些用户名和密码非常像是用于机场和其他重要设施的，而英国政府官员大概在 2017 年 4 月 1 日关于针对机场和核设施的“实际存在的网络威胁”做出过警告。哎呀。
* 这种更加激进的排查还引起了民间安全研究者的注意，安全公司 Radware 在 2017 年 4 月 6 日发表了一篇关于我的项目的文章。这个公司把它叫做“BrickerBot”。显然，如果我要继续增加我的物联网防御措施的规模，我必须想出更好的网络映射与检测方法来应对蜜罐或者其他有风险的目标。
* 2017 年 4 月 11 日左右的时候发生了一件非常不寻常的事情。一开始这看上去和许多其他的互联网服务提供商下架事件相似，一个叫 Sierra Tel 的半本地的互联网服务提供商在一些 Zyxel 设备上使用了默认的 telnet 用户名密码 supervisor/zyad1234。一个 Mirai 使用者发现了这些有漏洞的设备，而我的僵尸网络紧随其后，2017年精彩绝伦的 BCM CLI 战争又开启了新的一场战斗。这场战斗并没有持续很久。它本来会和 2017 年的其他数百起互联网服务提供商下架事件一样，如果不是在尘埃落定之后发生的那件非常不寻常的事情的话。令人惊奇的是，这家互联网服务提供商并没有试着把这次网络中断掩盖成某种网络故障、电力超额或错误的固件升级。他们完全没有对客户说谎。相反，他们很快发表了新闻公告，说他们的调制解调器有漏洞，这让他们的客户们得以评估自己可能遭受的风险。这家全世界最诚实的互联网服务提供商为他们值得赞扬的公开行为而收获了什么呢？悲哀的是，它得到的只是批评和不好的名声。这依然是我记忆中最令人沮丧的“为什么我们得不到好东西”的例子，这很有可能也是为什么 99% 的安全错误都被掩盖而真正的受害者被蒙在鼓里的最主要原因。太多时候，“有责任心的信息公开”会直接变成“粉饰太平”的委婉说法。
* 在 2017 年 4 月 14 日，国土安全部关于“BrickerBot 对物联网的威胁”做出了警告，我自己的政府把我作为一个网络威胁这件事让我觉得他们很不公平而且目光短浅。跟我相比，对美国人民威胁最大的难道不应该是那些部署缺乏安全性的网络设备的提供商和贩卖不成熟的安全方案的物联网设备制造商吗？如果没有我，数以百万计的人们可能还在用被入侵的设备和网络来处理银行业务和其他需要保密的交易。如果国土安全部里有人读到这篇文章，我强烈建议你重新考虑一下保护国家和公民究竟是什么意思。
* 在 2017 年 4 月底，我花了一些时间改进我的 TR069/64 攻击方法，然后在 2017 年 5 月初，一个叫 Wordfence 的公司（现在叫 Defiant）报道称一个曾给 Wordpress 网站造成威胁的基于入侵 TR069 的僵尸网络很明显地衰减了。值得注意的是，同一个僵尸网络在几星期后使用了一个不同的入侵方式暂时回归了（不过这最终也同样被化解了）。
* 在 2017 年 5 月，主机公司 Akamai 在它的 2017 年第一季度互联网现状报告中写道，相比于 2016 年第一季度，大型（超过 100 Gbps）DDoS 攻击数减少了 89%，而总体 DDoS 攻击数减少了 30%。鉴于大型 DDoS 攻击是 Mirai 的主要手段，我觉得这给这些月来在物联网领域的辛苦劳动提供了实际的支持。
* 在夏天我持续地改进我的入侵技术军火库，然后在 7 月底我针对亚太互联网络信息中心的互联网服务提供商进行了一些测试。测试结果非常令人吃惊。造成的影响之一是数十万的 BSNL 和 MTNL 调制解调器被禁用，而这次中断事故在印度成为了头条新闻。考虑到当时在印度和中国之间持续升级的地缘政治压力，我觉得这个事故有很大的风险会被归咎于中国所为，于是我很罕见地决定公开承认是我所做。Catalin，我很抱歉你在报道这条新闻之后突然被迫放的“两天的假期”。
* 在处理过亚太地区和非洲地区的互联网络信息中心之后，在 2017 年 8 月 9 日我又针对拉丁美洲与加勒比地区互联网络信息中心进行了大规模的清理，给这个大洲的许多提供商造成了问题。在数百万的 Movilnet 的手机用户失去连接之后，这次攻击在委内瑞拉被大幅报道。虽然我个人反对政府监管互联网，委内瑞拉的这次情况值得注意。许多拉美与加勒比地区的提供商与网络曾在我的僵尸网络的持续调节之下连续数个月逐渐衰弱，但委内瑞拉的提供商很快加强了他们的网络防护并确保了他们的网络设施的安全。我认为这是由于委内瑞拉相比于该地区的其他国家来说进行了更具侵入性的深度包检测。值得思考一下。
* F5 实验室在 2017 年 8 月发布了一个题为“狩猎物联网：僵尸物联网的崛起”的报告，研究者们在其中对近期 telnet 活动的平静表达了困惑。研究者们猜测这种活动的减少也许证实了一个或多个非常大型的网络武器正在成型（我想这大概确实是真的）。这篇报告是在我印象中对我的项目的规模最准确的评估，但神奇的是，这些研究者们什么都推断不出来，尽管他们把所有相关的线索都集中到了一页纸上。
* 2017 年 8 月，Akamai 的 2017 年第二季度互联网现状报告宣布这是三年以来首个该提供商没有发现任何大规模（超过 100 Gbps）攻击的季度，而且 DDoS 攻击总数相比 2017 年第一季度减少了 28%。这看上去给我的清理工作提供了更多的支持。这个出奇的好消息被主流媒体所完全忽视了，这些媒体有着“流血的才是好新闻”的心态，即使是在信息安全领域。这是我们为什么不能得到好东西的又一个原因。
* 在 CVE-2017-7921 和 7923 于 2017 年 9 月发表之后，我决定更密切地关注海康威视公司的设备，然后我惊恐地发现有一个黑客们还没有发现的方法可以用来入侵有漏洞的固件。于是我在 9 月中旬开启了一个全球范围的清理行动。超过一百万台硬盘录像机和摄像机（主要是海康威视和大华出品）在三周内被禁用，然后包括 IPVM.com 在内的媒体为这些攻击写了多篇报道。大华和海康威视在新闻公告中提到或暗示了这些攻击。大量的设备总算得到了固件升级。看到这次清理活动造成的困惑，我决定给这些闭路电视制造商写一篇[简短的总结][3]（请原谅在这个粘贴板网站上的不和谐的语言）。这令人震惊的有漏洞而且在关键的安全补丁发布之后仍然在线的设备数量应该能够唤醒所有人，让他们知道现今的物联网补丁升级过程有多么无力。
* 2017 年 9 月 28 日左右，Verisign 发表了报告称 2017 年第二季度的 DDoS 攻击数相比第一季度减少了 55%，而且攻击峰值大幅减少了 81%。
* 2017 年 11 月 23 日，CDN 供应商 Cloudflare 报道称“近几个月来，Cloudflare 看到试图用垃圾流量挤满我们的网络的简单进攻尝试有了大幅减少”。Cloudflare 推测这可能和他们的政策变化有一定关系，但这些减少也和物联网清理行动有着明显的重合。
* 2017 年 11 月底，Akamai 的 2017 年第三季度互联网现状报告称 DDoS 攻击数较前一季度小幅增加了 8%。虽然这相比 2016 年的第三季度已经减少了很多，但这次小幅上涨提醒我们危险仍然存在。
* 作为潜在危险的更进一步的提醒，一个叫做“Satori”的新的 Mirai 变种于 2017 年 11 月至 12 月开始冒头。这个僵尸网络仅仅通过一个 0-day 漏洞而达成的增长速度非常值得注意。这起事件凸显了互联网的危险现状，以及我们为什么距离大规模的事故只差一两起物联网入侵。当下一次威胁发生而且没人阻止的时候，会发生什么？Sinkholing 和其他的白帽或“合法”的缓解措施在 2018 年不会有效，就像它们在 2016 年也不曾有效一样。也许未来各国政府可以合作创建一个国际范围的反黑客特别部队来应对特别严重的会影响互联网存续的威胁，但我并不抱太大期望。
* 在年末出现了一些危言耸听的新闻报道，有关被一个被称作“Reaper”和“IoTroop”的新的僵尸网络。我知道你们中有些人最终会去嘲笑那些把它的规模估算为一两百万的人，但你们应该理解这些网络安全研究者们对网络上发生的事情以及不由他们掌控的硬件的事情都是一知半解。实际来说，研究者们不可能知道或甚至猜测到大部分有漏洞的设备在僵尸网络出现时已经被禁用了。给“Reaper”一两个新的未经处理的 0-day 漏洞的话，它就会变得和我们最担心的事情一样可怕。
### 3. 临别赠言
我很抱歉把你们留在这种境况当中，但我的人身安全受到的威胁已经不允许我再继续下去。我树了很多敌人。如果你想要帮忙，请看前面列举的该做的事情。祝你好运。
也会有人批评我，说我不负责任，但这完全找错了重点。真正的重点是如果一个像我一样没有黑客背景的人可以做我做到的事情，那么一个比我厉害的人可以在 2017 年对互联网做比这要可怕太多的事情。我并不是问题本身，我也不是来遵循任何人制定的规则的。我只是报信的人。你越早意识到这点越好。
-Dr Cyborkian 又名 janit0r，“病入膏肓”的设备的调节者。
--------------------------------------------------------------------------------
via：https://ghostbin.com/paste/q2vq2
作者：janit0r
译者：[yixunx](https://github.com/yixunx)
校对：[校对者ID](https://github.com/校对者ID)
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，
[Linux中国](https://linux.cn/) 荣誉推出
[1]:https://www.telecompaper.com/news/venezuelan-operators-hit-by-unprecedented-cyberattack--1208384
[2]:https://github.com/JeremyNGalloway/mod_plaintext.py
[3]:http://depastedihrn3jtw.onion.link/show.php?md5=62d1d87f67a8bf485d43a05ec32b1e6f