ZeroHour
Security Affairspublished ()ingested @securityaffairs

Muhstik Botnet Targeting Redis Servers Using Recently Disclosed Vulnerability

Vulnerabilities mentionedAll →

CVEVulnerabilityCVSSEPSSFlagsAffectedExposurePublished
CVE-2017-10271
Unauthenticated Remote Code Execution in Oracle WebLogic Server

Oracle WebLogic Server (CVE-2017-10271) contains a remote code execution flaw in its WLS Security component, where an XMLDecoder deserialization bug allows unauthenticated attackers to execute arbitrary code. The flaw is triggered by sending specially crafted XML requests to exposed WebLogic HTTP endpoints, so any instance whose WebLogic ports are reachable by untrusted users is at risk. Successful exploitation yields arbitrary code execution on the application server host, and the bug has been widely leveraged to install cryptominers and ransomware payloads. Organizations running Oracle WebLogic Server — particularly instances directly exposed to the internet — are affected, and the flaw has been on CISA's Known Exploited Vulnerabilities catalog since 2022-02-10 with confirmed ransomware use. Exploitation is assessed as essentially certain in the near term (EPSS 100.0%, 100th percentile), and defenders should treat it as an actively exploited, high-priority issue.

Do: Apply Oracle's updates for CVE-2017-10271 per Oracle's instructions (the vendor's required action for this KEV entry). Until patched, restrict network access to WebLogic HTTP/admin ports from untrusted networks and limit or remove access to the XML/WSAT endpoints used for exploitation. Because ransomware operators are known to exploit this flaw, review internet-exposed WebLogic instances for signs of compromise, such as unexpected processes, new scheduled tasks, webshells, or ransomware artifacts.

7.5100% KEV ransomware PoC ×3
  • Oracle WebLogic Server
large≈tens of thousands of internet-exposed WebLogic Server instances (total deployed install base likely higher, including internal-only servers)
CVE-2018-7600
Unauthenticated Remote Code Execution in Drupal Core (Drupalgeddon 2)

CVE-2018-7600, widely known as 'Drupalgeddon 2', is an unauthenticated remote code execution flaw in Drupal Core caused by insufficient input validation (CWE-20) in how the CMS processes certain structured request data. It can be triggered through multiple attack vectors, such as crafted parameters submitted to commonly used form and rendering features that are reachable by anonymous users with a single HTTP request. Successful exploitation lets an attacker run arbitrary code under the web application, typically resulting in complete site compromise, and CISA notes the flaw has been used in ransomware operations. Any site running unpatched Drupal 7.x or 8.x core is affected; Drupal's installed base at the time of disclosure was on the order of one million sites. Exploitation is confirmed in the wild (CISA KEV, added 2021-11-03), EPSS assigns a 100% probability of exploitation within 30 days, and no public PoC is recorded in the supplied data.

Do: Upgrade immediately per vendor instructions: Drupal 7.58, 8.5.1, or the corresponding 8.4.6/8.3.9 updates if you remain on older 8.x branches, prioritizing internet-facing sites. Because this flaw has been exploited in the wild and used in ransomware operations, also check patched sites for backdoors, unexpected administrator accounts, modified core files, and rotate credentials.

9.8100% KEV ransomware PoC ×4
  • Drupal Core Drupal 7.x prior to 7.58 and Drupal 8.x prior to 8.5.1 (prior to 8.4.6 on the 8.4.x branch and prior to 8.3.9 on the 8.3.x branch); version ranges per vendor ad
mass≈1,000,000 sites (Drupal's self-reported installed base at time of disclosure)
CVE-2019-2725
Unauthenticated Injection in Oracle WebLogic Web Services Enables RCE

CVE-2019-2725 is an easily exploitable injection flaw (CWE-74) in the Web Services subcomponent of Oracle WebLogic Server within Oracle Fusion Middleware, publicly documented as affecting WebLogic 10.3.6.0, 12.1.3.0, 12.2.1.1 and 12.2.1.3. It is triggered when an unauthenticated remote attacker sends attacker-controlled XML over HTTP to the WebLogic Web Services async response endpoint (the /_async/AsyncResponseService servlet), which processes the input unsafely. Successful attacks give the attacker takeover of the affected WebLogic server (remote code execution); in the 2019 exploitation wave this was used to install cryptocurrency miners and deploy ransomware. Any organization running affected Oracle WebLogic Server versions is exposed, with the greatest risk where the async/Web Services endpoints are reachable, especially on internet-facing servers. Exploitation is confirmed in the wild: CISA added the CVE to its KEV catalog on 2022-01-10 with ransomware use known and requires applying vendor updates, the EPSS probability of exploitation is 100% (100th percentile), and no public PoC is catalogued.

Do: Apply Oracle's updates per vendor instructions: this CVE was fixed by Oracle's April 2019 out-of-band WebLogic patch and is covered by subsequent Critical Patch Updates, so bring affected WebLogic 10.3.6.0/12.1.3.0/12.2.1.x servers to a patched level (CISA KEV requires this action). Until patched, restrict or remove the async response service deployments (wls9_async_response.war / wls_wsee_async_response.war, exposing the /_async/AsyncResponseService endpoint) and keep WebLogic ports off the internet. Also review logs for unsolicited POSTs to /_async/AsyncResponseService and for signs of dropped miners or ransomware payloads.

9.8100% KEV ransomware PoC
  • Oracle WebLogic Server (Oracle Fusion Middleware, Web Services subcomponent)
largetens of thousands of internet-exposed WebLogic servers (~30,000+ reported in 2019 internet scans), plus a much larger internal install base
CVE-2022-0543
Lua Sandbox Escape in Debian-packaged Redis Enables Unauthenticated RCE

CVE-2022-0543 is a Debian-specific Lua sandbox escape in the Redis key-value database, caused by a packaging issue in how the Debian-distributed redis-server build ships its embedded Lua interpreter (tracked as CWE-862). A remote, unauthenticated attacker can send a crafted Lua script through Redis' scripting interface (EVAL) to escape the Lua sandbox that normally limits scripts to safe operations. Successful exploitation yields remote code execution on the server in the context of the Redis process, giving the attacker control over stored data and the ability to run arbitrary system commands. Only Redis servers installed from Debian's (and derivative distributions') packages are affected; upstream Redis builds are not impacted. The flaw is actively exploited: CISA added it to the Known Exploited Vulnerabilities catalog on 2022-03-28, EPSS puts the 30-day exploitation probability at 99.4% (100th percentile), and botnet/worm activity against exposed Redis servers has been widely reported.

Do: Apply the updated redis-server packages published through Debian security updates (and equivalent updates in derivative distributions) per vendor instructions, restarting the Redis service after patching. As interim mitigation, restrict Redis exposure by binding to trusted interfaces only, enabling authentication, and firewalling port 6379 from untrusted networks. Check affected hosts for signs of compromise, such as unexpected Lua EVAL activity, unfamiliar child processes, or connections consistent with botnet/worm traffic.

10.099% KEV PoC
  • Debian (redis package; derived distributions affected via packaging) Redis persistent key-value database (Debian-specific builds)
masson the order of 100,000+ internet-exposed Redis servers, with Debian/Ubuntu-packaged installs a significant and plausibly large subset
Full article483 words · extracted from securityaffairs.com · click to collapse

The Muhstik botnet has been observed targeting Redis servers exploiting the recently disclosed CVE-2022-0543 vulnerability.

Muhstik is a botnet that is known to use web application exploits to compromise IoT devices, it has been around for at least 2018. Botnet operators monetize their efforts via XMRig combined with DDoS-for-hire services. 

The botnet leverages IRC servers for command-and-control (C2) communications, experts noticed that it has consistently used the same infrastructure since it first appeared in the threat landscape.

The bot propagates by compromising home routers, but experts observed multiple attempted exploits for Linux server propagation. The list of targeted routers include GPON home router, DD-WRT router, and the Tomato router

The bot includes exploits for Oracle WebLogic Server vulnerabilities CVE-2019-2725 and CVE-2017-10271, and the Drupal RCE flaw tracked as CVE-2018-7600.

Now researchers from Juniper Threat Labs observed the Muhstik botnet exploiting the CVE-2022-0543 Lua sandbox escape flaw that impacts Debian and Debian-derived Linux distributions.

“Juniper Threat Labs has uncovered an attack that targets Redis Servers using a recently disclosed vulnerability, namely CVE-2022-0543. This vulnerability exists in some Redis Debian packages. The attack started on March 11, 2022 from the same threat actor we’ve seen targeting confluence servers back in September 2021 and the same group targeting Log4j back in December.” reads the analysis published by Juniper. “The payload used is a variant of Muhstik bot that can be used to launch DDOS attacks.”

The vulnerability, which was rated 10 out of 10 for severity, could be exploited by a remote attacker with the ability to execute arbitrary Lua scripts to possibly escape the Lua sandbox and execute arbitrary code on the underlying machine.

The attacks exploiting the CVE-2022-0543 vulnerability started on March 11, 2022, the attackers were attempting to fetch the malicious script “russia.sh” using wget or curl from “106[.]246.224.219”. It saves it as “/tmp/russ” and executes it.

Experts pointed out that the vulnerability is not related to Redis, instead, it existed because the Lua library in some Debian/Ubuntu packages is provided as a dynamic library (Ubuntu Bionic and Trusty are not affected). Upon initializing the Lua interpreter, the “package” variable is automatically populated, and that in turn permitted access to arbitrary Lua functionality.  

To demonstrate this attack, the researchers set up a vulnerable Redis server and launched a proof of concept exploit code which is a Lua script that is using the “eval” command.
The researchers demonstrated that they are able to achieve code execution by dumping the contents of “/etc/passwd”.  

Muhstik botnet

The researchers shared Indicators of Compromise (IoCs) for these attacks.

“We advise those who may be vulnerable to patch their Redis service.” concludes the report. “Debian and Ubuntu have also released security advisories regarding this matter. Links are below: 

Follow me on Twitter: @securityaffairs and Facebook

[adrotate banner=”9″][adrotate banner=”12″]

Pierluigi Paganini

(SecurityAffairs – hacking, Muhstik botnet)

[adrotate banner=”5″]

[adrotate banner=”13″]



Text extracted automatically; images, tables and formatting may be missing. Original: https://securityaffairs.com/129549/cyber-crime/muhstik-botnet-targeting-redis-servers-using-recently-disclosed-vulnerability.html