惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

G
Google Developers Blog
博客园 - 三生石上(FineUI控件)
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
MongoDB | Blog
MongoDB | Blog
小众软件
小众软件
Y
Y Combinator Blog
博客园 - 聂微东
Google DeepMind News
Google DeepMind News
D
Docker
罗磊的独立博客
Microsoft Security Blog
Microsoft Security Blog
D
DataBreaches.Net
B
Blog
Vercel News
Vercel News
Recent Announcements
Recent Announcements
GbyAI
GbyAI
阮一峰的网络日志
阮一峰的网络日志
T
The Blog of Author Tim Ferriss
H
Hackread – Cybersecurity News, Data Breaches, AI and More
P
Proofpoint News Feed
酷 壳 – CoolShell
酷 壳 – CoolShell
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
Microsoft Azure Blog
Microsoft Azure Blog
宝玉的分享
宝玉的分享

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
Process Hollowing Detection: Your RAM is your treasure!
M.M · 2026-05-08 · via DEV Community

INTRODUCTION
Today, we are going to be diving into Process Hollowing, a cyber attack technique used by hackers to introduce and run malware into a victim machine.

In this article, you'll learn how it works, how to detect it on your system, and go through a practical carried out to detect Process Hollowing.

Before beginning the article, a huge thank you to HSC Consult for the mentorship efforts on this journey.

HOW DOES PROCESS HOLLOWING WORK?
On a very high level, this is just a bad untrusted process pretending to be a trusted process, in order to avoid detection systems. As a user, you could be seeing a notepad process running. Normal, right? Wrong! Someone made it look like a notepad process, but it really is malware running on your system.

Process Hollowing starts up a new program that is known and trusted, and right before the program executes, the attacker finds where the program code sits in memory, writes over the code with new code, and then executes the process. That way, when the trusted program fully executes, it executes the injected malware.

Just as an example, let's say that the trusted process is notepad.exe, a tool used very frequently on Windows system.

HOW RAM CAN BE USEFUL IN DETECTING PROCESS HOLLOWING
When this happens, memory analysis is very helpful in detecting this technique. Note that Process Hollowing can be classified as 'Living-off-the-land' (LOTL) attack, since it makes use of legitimate processes (Read more on LOTL here).

To understand how RAM is useful here, let's review how RAM works.

RAM is the short-term working memory of the computer. What does that even mean? Well, think about this analogy:

  • Say you wanted to have breakfast in the morning. You think "Let me warm some bacon and toast on the pan", and then "The heat is too much... there, much better!", "All done. It's smells so good", "Wow, that was delicious, I was so hungry!"
  • Once you're done with breakfast, you leave the pan and plate in the sink to clean later.
  • If I walked into your kitchen and find the pan and plate in the sink, that will be enough evidence to tell me that you ate breakfast today.

Well, sometimes when attacks happen, there exists the typical evidence that tells us that someone did something, just like the dishes in the sink. This often appears as malicious files within the computer storage. But what if they don't leave a file behind? What if they clean up the sink after breakfast? In that case, we can search the memory for 'thoughts', such as the thoughts you had while having bacon and toast for breakfast. Those thoughts tell us what you did sequentially, what you ate, when you ate it and that you actually enjoyed breakfast. RAM works the same way; it logs the processes within the computer as they are happening, regardless of whether there is evidence on the disk or not.

One thing to remember about RAM is that it is volatile (can be lost very easily). RAM requires continuous power to maintain data. During a forensic investigation, never switch off the computer! Capture RAM first, before the computer is powered off. If the power goes off, you lose your treasure.

PRACTICAL
The Volatility is a supreme tool for memory analysis! Once you have the captured RAM, you can generate a couple of reports using the tool, which you'll analyse for the threat landscape.

For this practical, 3 reports were analysed:

  1. pstree_output.txt (Process Tree Analysis)
  2. malfind_results.txt (Memory Injection/In-memory artifacts)
  3. netscan_results.txt (Network Connection artifacts)

Analysis of Parent-Child Process Hierarchy
The pstree_output.txt report maps out all running processes and their parent-child relationships, helping you spot processes spawned by unusual or unexpected parents.

The goal here is to confirm that each active process came from a valid parent in order to create a baseline of system integrity. Every entry was compared to the Parent Process ID (PPID) standard Windows architectural expectations.

Process ..... PID .... Parent Process .... PPID .... Is this normal?
svchost.exe .... 720 .... services.exe .... 512 .... Yes
svchost.exe .... 884 .... services.exe .... 512 .... Yes
svchost.exe .... 3880 .... notepad.exe .... 2250 .... No!

svchost.exe is a standard Windows system process, completely legit. It seems that the third svchost.exe was spawned by the parent process notepad.exe, which is very unusual and suspicious.

Analysis of the Network Connections
The detected PID 3880 was cross-analyzed using grep CLI command across the remaining forensic datasets. The results showed one entry each within the other reports.

The netscan_results.txt report lists all active and recently closed network connections, showing which processes are communicating externally and on what ports.

PID 3880 in the networkscan file indicates that the process is communicating outside of the local network and is linked to an active, established connection to an external host. Looking at the report, there was an external IP connection, with Port of destination: 443 (HTTPS) and State of Connection: Established.

Using Port 443 enables the malicious communication to evade perimeter firewall inspections and interact with legitimate online traffic. This established relationship demonstrates that the process is actively controlled by an outside party rather than just being inactive, Command & Control (C2) in action!

Malware Analysis and Memory Artefacts
The malfind_results.txt report scans memory regions for suspicious executable code, flagging areas with RWX permissions or embedded PE headers that suggest injection.

svchost.exe (PID 3880) showed an indicator of a malicious payload; it showed indicators of injected shellcode or a reflective PE (Portable Executable) loader in the existence of RWX memory. This means that the file had read, write and execute permissions, which is very unusual. PAGE_EXECUTE_READ or PAGE_READONLY is the architectural standard for a common system component such as svchost.exe.

Magic Bytes were found at the beginning of the memory section after more examination of the PID 3880 raw hexadecimal data:
Header in Hex: 4d 5a
Translation of ASCII: MZ

These bytes are the Windows Portable Executable (.exe) Magic Header, a raw executable header residing unbacked in the memory space of the running process, indicating that a malicious binary has been loaded straight into the RAM of PID 3880.

CONCLUSION
Process Hollowing is a sneaky technique, but it's not unbeatable. As we saw in the practical, memory analysis using Volatility can expose the attack through a combination of clues — an unusual parent-child process relationship, an active C2 connection, and RWX memory regions containing a raw PE header. You can use Volatility to get more clues! No single indicator tells the full story, but together they paint a clear picture of compromise. The takeaway? Always capture RAM before pulling the plug, and let the memory do the talking. The attacker may have skipped leaving files on disk, but the "thoughts" in RAM gave them away.