Security researchers have identified a new type of cyber attack causes damage to Internet of Things (IoT) devices, rather than ensnaring them into a botnet.
Dubbed Permanent Denial-of-Service (PDoS), the attacks can be highly damaging, resulting in the need to replace or reinstall hardware, researchers explain: security flaws are abused to destroy the firmware and/or basic functions of system.
One of the tools used to launch such attacks is called BrickerBot, and Radware researchers observed two variants starting March 20, 2017. One of them, however, had a short life and remains inactive, while the other continues to operate. Both, however, have had the same purpose: to compromise IoT devices and corrupt their storage.
Both bots started PDoS attempts on the same date and they were discovered within one hour of each other. However, while the first showed intense activity over its short life, the second displayed lower intensity, but has been more thorough in its attacks and has been also concealing its location using TOR (The Onion Router) egress nodes.
To compromise devices, BrickerBot uses Telnet brute force, a method previously associated with the Mirai botnet, which abused infected devices to launch distributed denial of service (DDoS) attacks.
Once it has successfully accessed a device, the PDoS bot performs a series of Linux commands meant to ultimately corrupt storage. Next, it also attempts to disrupt Internet connectivity and device performance, and to wipe all files on the device.
“Among the special devices targeted are /dev/mtd (Memory Technology Device – a special device type to match flash characteristics) and /dev/mmc (MultiMediaCard – a special device type that matches memory card standard, a solid-state storage medium),” Radware researchers reveal.
The attack is targeted specifically at Linux/BusyBox-based IoT devices that have the Telnet port open and exposed publically on the Internet. These are the same type of devices that Mirai and related IoT botnets have been targeting.
The recorded PDoS attempts originated from a limited number of IP addresses worldwide, with all devices exposing port 22 (SSH) and running an older version of the Dropbear SSH server. These were identified as Ubiquiti network devices.
The security researchers also identified a second type of PDoS attempts, with a different command signature, which hid their source IP addresses behind TOR nodes. Still ongoing, these attacks attempt to brute-force the Telnet login using the root/root and root/vizxv username-password pairs, use more thorough commands, and target a much broader range of storage devices.
These attacks don’t use ‘busybox’ but attempt both ‘dd’ and ‘cat,’ whichever is available on the breached device, the researchers say. In the end, these attacks also attempt to remove the default gateway, wipe devices, and disable TCP timestamps. With the help of extra commands, the attackers attempt to flush all iptables firewall and NAT rules and add a rule to drop all outgoing packets.
Related: Mirai Switches to Tor Domains to Improve Resilience
Related: New “Ghost Host” Technique Boosts Botnet Resiliency