Soluzioni Corporate di backup Alessandro Stesi

Corporate backup solutions Self-protection tests

Introduction

Corporate backup solutions: In light of the growing number of ransomware attacks in which cryptolockers stop database processes to unlock database files for encryption (Cerber, GlobeImposter, Rapid, Serpent) and they can encrypt local and network backups to request a ransom payment (Rapid, Spora), we decided to test the self-protection capabilities of the best backup solutions used in the corporate environments available for testing.

The test aims to check the sustainability of the processes and services of the products against the typical attacks on the security software described below, as well as the self-protection of the local backup and of the product files. Ransomware can encrypt configuration files and local backup files that belong to a backup program by disabling file recovery. In addition, once access to the agent or server processes is obtained, the attacker can delete backup copies of files not only locally, but also in the cloud on behalf of a backup solution.

This document is a summary of the enterprise backup solutions test report and includes a description of the test environment, a list of tested solutions and their versions, an overview of test scenarios, as well as results and conclusions based on these results. We do not classify the tested solutions and do not award any prize, but we provide the results “as is” for informational purposes only.

 

Testing environment

The tests were conducted on the virtual machines of:

 – Windows 8.1 SP1 32 bit build 9600

 – Windows 10 Enterprise 64 bit Build 16299

 – Windows Server 2012 R2 Standard 64 bit v. 6.3.9600 Build 9600

We have tested backup solutions on platforms
32 and 64 bit because the process injection techniques used in the test scenarios differ on these platforms. In addition, 32 and 64 bit product builds may contain a different set of features, including self-protection, and their implementation may depend on the architecture of the operating system.

 

Tested products

Corporate backup solutions: The most recent versions of the following products available at the time of testing were tested:

Product name

Components

Version

Management Server

Agent

12.5 9010

12.5 9010

Unified Data Protection Server

Unified Data Protection Client

6.5.4175 Aggiornamento 2 Build 667

6.5.4175.791 v.r6.5

Backup & Replication

Agent for Microsoft Windows

9.5 Aggiornamento 3

2.1.0.423

Veritas Backup Exec

Server

Agent Utility pour Windows

16.0 Rev. 1142

16.0 Rev. 1142-1632

 

Each product was installed with the default settings and updated before running the test.

 

Corporate backup solutions – Test scenarios

The test suite includes 31 tests that simulate attacks on local backup files, product files, processes, services and cloud storage that aim to block the backup and restore service. The “Product File Protection” test category contains simple tests aimed at destroying backup and application files making it impossible to recover data encrypted by ransomware.

The second group of tests “Protection of processes and products services” is essential for self-protection since the malware can inject the malicious code into a backup agent and act on behalf of a backup solution obtaining all the necessary privileges to check the backup files. At the behest of an attacker, a malicious process can interrupt processes and services with consequent blocking of the backup and restore application

or deleting backup files on behalf of a backup solution. The latest series of tests is “Cloud backup and recovery protection” and is addressed to communication interfaces with cloud storage. The attack of DNS poisoning or the improper use of the CLI can cause the cloud backup service to be interrupted.

 

Corporate backup solutions – Conclusions

The purpose of the test was to verify the backup software’s self-protection capabilities to protect related files, processes, services and cloud storage from scenarios that can potentially be run by ransomware.

The results showed that most of the tested products are not ready in many cases to counter ransomware-type attacks by allowing a potential attacker to block user backups and disable backup and restore services. Only Acronis Backup has shown good results with an 87% and 81% effectiveness rate for 32-bit and 64-bit products, similarly providing complete self-protection and service sustainability features.

Download the complete department here: Nio Guard Independent Study_IT

 

[btnsx id=”2931″]

Related articles:

Modern Data Protection: The Difference is in the Cloud

BaaS | Acronis Cloud Backup

BaaS | Cloud Backup | Backup as a Service

[Name change] Acronis Data Cloud –> Acronis Cyber Cloud

Stay in control of your fast-moving, quick-shifting data

A Hybrid Cloud Backup Solution for System Integrator and reseller

Share


RSS

More Articles…

Categories …

Tags

RSS Unknown Feed

RSS Full Disclosure

  • User Enumeration in IServ Schoolserver Web Login September 11, 2025
    Posted by naphthalin via Fulldisclosure on Sep 10“I know where your children go to school.” The web front end of the IServ school server from IServ GmbH allows user enumeration. Responses during failed login attempts differ, depending on if the user account exists, does not exist and other conditions. While this does not pose a […]
  • Re: Apple’s A17 Pro Chip: Critical Flaw Causes Dual Subsystem Failure & Forensic Log Loss September 11, 2025
    Posted by Matthew Fernandez on Sep 10Can you elaborate on why you consider this high severity? From the description, it sounds as if this behaviour is fail-closed. That is, the effects are limited to DoS, with security properties preserved.
  • Defense in depth -- the Microsoft way (part 92): more stupid blunders of Windows' File Explorer September 8, 2025
    Posted by Stefan Kanthak via Fulldisclosure on Sep 08Hi @ll, this extends the two previous posts titled Defense in depth -- the Microsoft way (part 90): "Digital Signature" property sheet missing without "Read Extended Attributes" access permission and Defense in depth -- the Microsoft way (part 91): yet another 30 year old bug of the […]
  • Critical Security Report – Remote Code Execution via Persistent Discord WebRTC Automation September 8, 2025
    Posted by Taylor Newsome on Sep 08Reporter: [Taylor Christian Newsome / SleepRaps () gmail com] Date: [8/21/2025] Target: Discord WebRTC / Voice Gateway API Severity: Critical 1. Executive Summary A proof-of-concept (PersistentRTC) demonstrates remote code execution (RCE) capability against Discord users. The PoC enables Arbitrary JavaScript execution in a victim’s browser context via WebRTC automation. […]
  • Submission of Critical Firmware Parameters – PCIe HCA Cards September 8, 2025
    Posted by Taylor Newsome on Sep 08*To:* support () mellanox com, networking-support () nvidia com *From:* Taylor Christian Newsome *Date:* August 20, 2025 *Dear Mellanox/NVIDIA Networking Support Team,* I am writing to formally submit the critical firmware parameters for Mellanox PCI Express Host Channel Adapter (HCA) cards, as detailed in the official documentation available here: […]
  • SEC Consult SA-20250908-0 :: NFC Card Vulnerability Exploitation Leading to Free Top-Up in KioSoft "Stored Value" Unattended Payment Solution (Mifare) September 8, 2025
    Posted by SEC Consult Vulnerability Lab via Fulldisclosure on Sep 08SEC Consult Vulnerability Lab Security Advisory < 20250908-0 > ======================================================================= title: NFC Card Vulnerability Exploitation Leading to Free Top-Up product: KioSoft "Stored Value" Unattended Payment Solution (Mifare) vulnerable version: Current firmware/hardware as of Q2/2025 fixed version: No version numbers available CVE number:...
  • FFmpeg 7.0+ Integer Overflow in FFmpeg cache: Protocol (CacheEntry::size) September 8, 2025
    Posted by Ron E on Sep 08An integer overflow vulnerability exists in the FFmpeg cache: URL protocol implementation. The CacheEntry structure uses a 32-bit signed integer to store cache entry sizes (int size), but the cache layer can accumulate cached data exceeding 2 GB. Once entry->size grows beyond INT_MAX and new data is appended, an […]
  • FFmpeg 7.0+ Integer Overflow in DSCP Option Handling of FFmpeg UDP Protocol September 8, 2025
    Posted by Ron E on Sep 08A vulnerability exists in the FFmpeg UDP protocol implementation ( libavformat/udp.c) where the dscp parameter is parsed from a URI and left-shifted without bounds checking. Supplying a maximum 32-bit signed integer (2147483647) triggers undefined behavior due to a left shift that exceeds the representable range of int. This results […]
  • FFmpeg 7.0+ Integer Overflow in UDP Protocol Handler (fifo_size option) September 8, 2025
    Posted by Ron E on Sep 08A signed integer overflow exists in FFmpeg’s udp.c implementation when parsing the fifo_size option from a user-supplied UDP URL. The overflow occurs during multiplication, which is used to compute the size of the circular receive buffer. This can result in undefined behavior, allocation failures, or potentially memory corruption depending […]
  • FFmpeg 7.0+ LADSPA Filter Arbitrary Shared Object Loading via Unsanitized Environment Variables September 8, 2025
    Posted by Ron E on Sep 08The ladspa audio filter implementation (libavfilter/af_ladspa.c) in FFmpeg allows unsanitized environment variables to influence dynamic library loading. Specifically, the filter uses getenv("LADSPA_PATH") and getenv("HOME") when resolving the plugin shared object (.so) name provided through the file option. These values are concatenated into a filesystem path and passed directly into […]

Customers

Newsletter

{subscription_form_1}