DDoSers are abusing the Plex Media Server to make attacks more potent

Distributed denial-of-service attackers have seized on a new vector for amplifying the junk traffic they lob at targets to take them offline: end users or networks using the Plex Media Server.

DDoS amplification is a technique that leverages the resources of an intermediary to increase the firepower of attacks. Rather than sending data directly to the server being targeted, machines participating in an attack first send the data to a third party in the form of a request for a certain service. The third party then responds with a much larger payload to the site the attackers want to take down.

So-called amplification attacks work by sending the third parties requests that are manipulated so they appear to have come from the target. When the third parties respond, the replies go to the target rather than the attacker device that sent the request. One of the most powerful amplifiers used in the past was the memcached database caching system, which can magnify payloads by a factor of 51,000. Other amplifiers include misconfigured DNS servers and the Network Time Protocol, to name only three.

On Thursday, DDoS mitigation service Netscout said that DDoS-for-hire services recently turned to misconfigured Plex Media Servers to amplify their attacks. The Plex Media Server is software that lets people access the music, pictures, and videos they store on one device with other compatible devices. The software runs on Windows, macOS, and Linux.

In some cases—such as when the server uses the Simple Service Discovery Protocol to locate universal plug-and-play gateways on end users’ broadband modems—the Plex service registration responder gets exposed to the general Internet. Responses range from 52 bytes to 281 bytes, providing an average amplification factor of about 5.

Netscout said that it has identified about 27,000 servers on the Internet that can be abused this way. To differentiate from plain-vanilla, generic Simple Service Discovery Protocol amplification DDoSes, the company is referring to the new technique as Plex Media SSDP or PMSSDP.

“The collateral impact of PMSSDP reflection/amplification attacks is potentially significant for broadband Internet access operators whose customers have inadvertently exposed PMSSDP reflectors/amplifiers to the Internet,” Netscout researchers Roland Dobbins and Steinthor Bjarnason wrote. “This may include partial or full interruption of end-customer broadband internet access, as well as additional service disruption due to access/distribution/aggregation/core/peering/transit link capacity consumption.”

In a statement, a Plex spokeswoman wrote:

The researchers who reported on this issue did not provide any prior disclosure, but Plex is now aware of the problem and is actively working on addressing it. This issue appears to be limited to a small number of media server owners who have misconfigured their firewalls by allowing UDP traffic on device-discovery ports from the public internet to reach their servers, and our current understanding is that it does not allow an attacker to compromise any Plex user’s device security or privacy. Plex is testing a simple patch that adds an extra layer of protection for those servers that may have been accidentally exposed and will release it shortly.

The researchers said that wholesale filtering of UDP data over port 32414 by network operators (not end users) has the potential to block some legitimate traffic. Instead, the researchers said operators (again, not end users) should identify PMSSDP nodes on their network that can be abused as DDoS reflectors or amplifiers. The researchers also recommended that ISPs disable SSDP by default in the equipment they provide to subscribers.

The forums section at Plex.tv provides these two threads that end users can peruse to best address the issue.

Post updated to add the third-to-last and last paragraphs.