System administrators have long been wary of the security implications of Simple Network Management Protocol (SNMP), but a recent experiment by "ethical hacking" group GNUCitizen has shown that many SNMP-enabled devices are left unguarded and may be prone to giving away sensitive information.

In a random scan of 2.5 million IP addresses via SNMP, the group found that many devices gave away names, models and in some cases the patch state of the OS.

SNMP has a number of known security weaknesses, including the fact that it is susceptible to brute force and dictionary attacks, and is typically accessed over User Datagram Protocol (UDP), which is vulnerable to IP spoofing.

The use of UDP means that large numbers of systems can be scanned rapidly via SNMP, in a shorter time than with TCP-based protocols, according to GNUCitizen researcher Adrian Pastor.

Because of SNMP's weaknesses, gaining read-only access to such devices over the Internet is a potentially serious security problem, according to Pastor.

"Even if a cracker only gained read access to a device/server via a SNMP community string, sometimes it would be possible to extract sensitive information such as usernames and passwords which would eventually lead to a compromise of the targeted systems," he said in a post on 3 March 2008. "In order to accomplish this, all that is needed by the attacker is knowledge of an interesting OID to query. My point is that SNMP read access could a enough to fully own a device."

GNUCitizen found that out of the 2.5 million random IP addresses scanned, 5,320 addresses responded to the submitted SNMP requests.

The low response rate is partly due to the fact that SNMP is supported mainly by embedded devices such as routers, which make up a small proportion of all IP addresses, Pastor said.

The top type of device responding was the Arris Touchstone Telephony Modem, a VoIP modem, which accounted for more than 35 percent of devices discovered.

Other common devices discovered included Cisco routers, Apple AirPort wireless base stations, ZyXel Prestige routers, Netopia routers and Windows 2000 servers.

These responses alone could be interesting for research, Pastor said. "For instance, if researching remote SNMP vulnerabilities, it would make sense to focus on a type of device that is widely spread through the Internet," he wrote.

Interesting information yielded up included a Windows server giving a full list of usernames, a BT Voyager 2000 router leaking ISP credentials including password, a HP JetDirect printer leaking the administration password and Dynamic DNS credentials disclosed by ZyXel Prestige routers.

"Lots of devices leak way too much information via SNMP read access," Pastor said on the group's website.

The research isn't groundbreaking, the group admitted, but rather is meant to "get an idea of the current state of remote SNMP hacking", Pastor stated.

Among others, the SANS Institute has long warned about the dangers of SNMP, warning in a typical document that "sniffed SNMP traffic can reveal a great deal about the structure of your network, as well as the systems and devices attached to it," and that intruders can use the information to plan attacks.

SANS advises administrators to disable SNMP unless it is absolutely required, or alternatively to take care in configuration.