[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
▼ Settings Mobile Home
/g/ - Technology

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • You may highlight syntax and preserve whitespace by using [code] tags.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


File: 1794936719-img_7875.jpg (493 KB, 3904x2171)
493 KB JPG
Two critical vulnerabilities discovered in nginx:

CVE-2026-123007
>This issue affects NGINX Plus or NGINX Open Source configurations that use HTTP/2. When NGINX processes undisclosed traffic, the flow control window can exceed the allowed threshold, potentially resulting in a heap buffer overflow typically followed by an NGINX worker segmentation fault. Additionally, attackers can execute code on systems with Address Space Layout Randomization (ASLR) disabled or when the attacker can bypass ASLR.
>Impact: An attacker could leverage this vulnerability to crash the NGINX worker process, corrupt its memory, or potentially execute code remotely. There is no control plane exposure; this is a data plane issue only.

CVE-2026-123633
>When NGINX Plus or NGINX Open Source is configured to use HTTP/2 for client connections and proxy client requests to a backend, a specially crafted request can result in use-after-free in the worker process, if this process is running low on worker_connections. Additionally, attackers can execute code on systems with Address Space Layout Randomization (ASLR) disabled or when the attacker can bypass ASLR.

Needless to say, any Rust-based solution didn't have these issues.

>but HURR DURR .unwrap()
lol
>>
File: images (10).jpg (8 KB, 225x225)
8 KB JPG
>http2
>>
>>110368210
Another vibeGOD victory
Based vibeGODS finding vulnerabilities
>>
>>110368210
>if this process is running low on worker_connections
good luck getting there on my most budget of vps's
>>
Rust chads won again. Imagine still trusting anything written in C to run as root of all things. lol lmao.
>>
File: majo rika face.jpg (79 KB, 640x480)
79 KB JPG
>>110368237
>someone pushes a fix
>clud run a diff on the patch
>AI CHADS DID THIS!!!!!
>>
>>110368296
>pays monthly fee to a jew
>"I'm a CHAD"
>>
i also have http3 enabled for the full rce experience
>>
>>110368210
tell me one webservice that uses HTTP/2

https://nginx.org/en/docs/http/ngx_http_v2_module.html
>This module is not built by default, it should be enabled with the --with-http_v2_module configuration parameter.

Total fucking nothingburger, also kinda explains why there would be holes in it as why pay attention to piece no one fucking uses.
>>
>>110368341
You have 0 experience in this don't you? HTTP/2 comes prebuilt into nginx on every major distro. Of course it's up to you to enable it with "http2 on;" but it's present in the binary and a lot of websites use it.
>>
File: 4.png (216 KB, 1167x1689)
216 KB PNG
>>110368341
>nothingburger
>well, no one uses it like that
>i'm not crying
>>
>>110368341
It's enabled on every single Nginx VM we turn up in our fleet, because of multiplexing/binary framing over a single connection. Or it was turned on until this morning, anyway.

Also, enabled by default for f5's ingress controller version, so you can bet there's some butthurt on cloud providers this morning.
>>
>>110368341
bro's still in 2015
>>
>>110368341
>This module is not built by default
this just means that, if you build nginx from source, you need to provide that flag to include the module. but most binary packages include it.
>>
>>110368458
It's also enabled/disabled at the vhost level. Pour one out for guys forced to docomuent compliance across multiple domains soread out across multiple VMs/containers.
>>
>>110368465
Nothing a grep/sed command can't fix. Also if you're not using global confs that get imported to every server block you're not gonna make it.
>>
>>110368341
It's not the ignorance of some posters on /g/ that blows my mind. It's how certain they are that they know what they're talking about.

If anonymity brings out your true self, fuck man...
>>
>>110368210
Nothingburger, as usual. Kek kill yourself
>>
>>110368225
Agreed, migrate everyone to HTTP/3 immediately.
>>
>>110368210
sounds a lot like a nothingburger
>>
>>110368576
Problem with that is it's slow. I think Nginx has some clever eBPF stuff to make it fast but fuck knows how to configure that?
>>
>>110368646
>call new protocol QUIC
>it's slow
What did they mean by this?
>>
>>110368646
>slow
?
i only know that it's being blocked in 3rd world
>>
>>110368658
>>110368663
It's slow because TCP is in the kernel, QUIC uses UDP which runs in userspace (more overhead). It doesn't mean it can't be made fast though.
>>
>>110368210
want a burger but all I have is buns
>>
>>110368576
I tried, but when I asked our security teams if they would open UPD port 443 they lost their fucking minds, so I dropped it. Jesus, it's just a different protocol, guys.
>>
>>110368672
I mean, half the people wringing their hands over Linux these days bitch about the monolithic kernel anyway, so...
>>
>>110368672
Can you just io_uring for UDP though? then you get kernel speed
>>
>>110368210
nothingburger status?
>>
>>110368673
Think outside the bun.
>>
>>110368699
No idea. Need to do more research on how to actually configure Nginx to be fast. Out of the box it will perform like dogshit though if you just
http3 on;
.

It's good enough for small / static files though.
>>
>>110368210
>Leveraging CVE-2026-123007 for RCE requires a significant amount of network traffic to be transmitted, approximately 128 GB of HTTP/2 data. Both available network bandwidth and connection timeouts will place practical limitations on Internet exploitation.
>>
>>110368210
>CVE-2026-123633 needs the server under pressure. The request must open an outbound connection, through any *_pass directive or a resolver, while the attacker holds at least 15/16 of worker_connections, about 960 of 1,024 by default.
>>
web servers become obsolete, same as their CVEs..

runtime should run the web server, not the web server

>>110368576
>Agreed, migrate everyone to HTTP/3 immediately.

youre outdoooted, migrate to http/4 or better, send the -ACK packet with self.

--
i would use http/1.1 but curl dude took corpo money and deleted pipelining. so now http/2 is what allows pipelining renamed to multiplexing.
>>
>>110368769
>runtime should run the web server, not the web server
That's what proxy_protocol, etc, is for.

Web server should serve the web still
App server should do backend stuff
>>
>>110368733
>Finally, the exploitation strategy developed depends on a specific x86-64 glibc EVEX/YMM memcpy implementation selected at runtime, based on the avx512f and avx512vl CPU features. Alternative RCE exploitation strategies may be viable, but none are currently known.
>>
>>110368769
>runtime should run the web server
I will still not run EMCAScript on the server.

To be serious though, SNI, TLS certs, redirects, and all the typical web layer bullshit mean you're just moving Nginx from a web server to the reverse proxy layer anyway.
>>
>>110368210
another day, another fake bullshit CVE that's impossible to abuse

>muh .unwrap()
unwrap did actually cause massive real world damages. Not a single person on this world will be affected by this cybersecurity scam.
>>
>>110368809
We'll see. I remember people sure that php2shell would never have an exploit, and then someone FAFOed courtesy of China state actors.
>>
>>110368210
I am still waiting to get hacked by that big bad php vulnerability.
Just like 70% of servers running php out there.
Still didn't update, because i deemed it unneccessary. So far i am correct.
>>
>>110368821
>someone FAFOed
aka they moved the goalposts 10 times and downgraded security until the exploit worked
>>
>>110368821
Damages inflicted by unwrap: Billions $$$

Damages inflicted by a php vulnerability that was only ever able to be replicated in controlled environments specifically rigged to get exploited: 0 $
>>
>>110368831
I don't think Palo Alto let their clients firewall rules get compromised, but hey, I could be wrong.
>>
>>110368850
>I don't think Palo Alto let their clients firewall rules get compromised
I mean they've hard-coded root SSH passwords and keys before, so I wouldn't put it past them
>>
>>110368850
do you really beleive government propaganda when they shout MUH CHYNA BAD?
Where you also outraged and concerned about the big bad weather balloon, when your government told you that this Chinese balloon has nuclear bombs stacked to attack the Free Western World?
>>
>>110368846
I'm not disputing that unwrap or left shift or log4j was worse. Calm down. I am saying multiple Taiwanese and Korean orgs got their iptables and Active Directories lifted off their Pan-OS devices due to php2shell.

>>110368858
I'm sorry, what?
>>
>>110368866
and i am saying that you are lying
>>
>>110368877
Okay, champ. You can look it up.
>>
>>110368890
what can i look up?
The French governments claim that Russian super hackers used advanced complex zero day exploits to hack their freezers... rather than some random kid discovering that they are open to the internet and their password is the current date?
>>
>>110368866
>I'm sorry, what?
https://security.paloaltonetworks.com/CVE-2019-15015
>>
File: file.png (1.99 MB, 1023x1024)
1.99 MB PNG
>>
>>110368890
I'm not looking up shit. Post proof yourself you worthless faggot.
Seriously what is this fucking type of arguing? Why would I look it up when you're making the claim? Worthless retard
>>
>>110368904
Well who else would benefit from spoiling food supplies and blood reserves in case of an open conflict
>>
>>110368904
Palo Alto published the vulnerability info themselves as part of a disclosure agreement. Sure, maybe the DoD put a gun to their head and said 'make your firewall devices look bad'. I can't help you there.

I will also say that companies I worked for had to post CSAF entries for government entities after an incident, and make them publicly available. But we didn't advertise where they were. The CSAF reports are still live though.
>>
>>110368957
Why do you imply benefit?
I would do it just for fun
>>
>>110368961
Your writing style makes you sound like you enjoy sucking cocks
>>
File: Jeordie.jpg (24 KB, 271x332)
24 KB JPG
>>110368933
this nigga look like balding Twiggy
>>
>>110368957
How does this even make sense!
If you are sitting on such a bomb, wouldnt you sray silent and just sit on it until there is a conflict?
The fact that Paris didn't get nuclear annihilated at the same time as their freezers dying proves you wrong.

You glowfags are truly the dumbest idiots you can find in the world
>>
>>110368941
>Post proof yourself you worthless faggot.
No, I don't think I will. I like the idea of you telling people "You know, I heard Palo Alto had their Pan-OS devices get exploited. I don't believe it, and you shouldn't either."
>>
>>110368957
If i find out that that some crucial french infrastructure is accessible with a known password, i would also abuse it and cause as much damage as possible.

I think the vast majority of people on the planet would.
>>
>>110368987
Nah most would install some form of ransomeware
>>
>>110368780
>Web server should serve the web still

web servers as a software doesnt make any sense anymore. this is a much much lesser stretch (a realistic one) than mid coders are replaced by SI and thus, coding and solved. you wont hear it because its too technical

>>110368789
>SNI, TLS certs, redirects, and all the typical web layer bullshit

ye, complicated bullshit. just run the runtime as a devops. its better from a developer perspective too. i mean from a good developer's perspective.

> Not as a whole—but traditional standalone web servers are losing ground at the edge. More sites now rely on CDNs, reverse proxies, and managed platforms (such as Cloudflare) to handle traffic before it reaches an origin server. W3Techs’ November 2026 survey lists Cloudflare Server at 34.6% of sites whose web server it identifies, ahead of Nginx at 29.7% and Apache at 20.5%. A site can use more than one server, so these figures overlap.

https://w3techs.com/technologies/overview/web_server
>>
>>110368210
I run nginx, without cloudflare in between, with HTTP2.

I do not care about your CVE. I dont feel like updating right now. There is a 99.999% chance that your CVE is impractical and unusable in the real world anyway.

No, i will not look into it.
No, i do not care about your /k/ope.
No, i am not arguing with some mentally ill idiot.
Get lost.
>>
>>110369027
For this, i would first have to develop such ransomware, which i wouldn't be assed to do.
Just trigger it and have some fun.
>>
caddychads where we at
>>
>>110369069
To get the passcodes you'd need the infrastructure to bruteforce it in the first place



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.