Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

Navigation
Guide That Contains This Content
[+] Expand All
[-] Collapse All

    Overview of Access Port Protection

    Port security features on QFX10000 switches can protect a switch against various types of attacks. Protection methods against some common attacks are:

    Mitigation of Ethernet Switching Table Overflow Attacks

    In an overflow attack on an Ethernet switching table, an intruder sends so many requests from new MAC addresses that the table cannot learn all the addresses. The attack forces the switch to send broadcast messages when it needs to send traffic to addresses for which it lacks MAC addresses. In addition to generating unnecessary traffic, the attacker might be able to sniff the broadcast packets.

    To mitigate such attacks, you can configure a limit for learned MAC addresses or allow only specific MAC addresses. Use the MAC limit feature to control the total number of MAC addresses that can be added to the Ethernet switching table for the specified interface or interfaces. By setting the MAC addresses that are explicitly allowed, you ensure that the addresses of network devices whose network access is critical are guaranteed to be included in the Ethernet switching table.

    Mitigation of Rogue DHCP Server Attacks

    By default, all access ports are untrusted, and all trunk ports are trusted with regard to DHCP. Trusted ports allow DHCP servers to provide IP addresses and other information to requesting devices. If someone connects an unauthorized DHCP server to a trusted port, the unauthorized server can start issuing IP addresses and configuration information to the network’s DHCP clients. The information provided to the clients by this server can disrupt their network access. The unauthorized server might also assign itself as the default gateway device for the network. An attacker can then sniff the network traffic and perpetrate a man-in-the-middle attack—that is, it misdirects traffic intended for a legitimate network device to a device of its choice.

    To mitigate this problem, set the interface to which the unauthorized server is connected as untrusted. That action blocks all ingress DHCP server messages from that interface. .

    Note: The switch logs all DHCP server packets that are received on untrusted ports. For example:

    5 untrusted DHCPOFFER received, interface xe-:0/0/2.0[65], vlan v1[10] server ip/mac 12.12.12.1/00:00:00:00:01:12 offer ip/client mac 12.12.12.253/00:AA:BB:CC:DD:01

    You can use these messages to detect unauthorized DHCP servers on the network.

    Note: If you attach a DHCP server to an access port, you must configure the port as trusted.

    Protection Against DHCP Starvation Attacks

    In a DHCP starvation attack, an attacker floods an Ethernet LAN with DHCP requests from spoofed (counterfeit) MAC addresses so that trusted DHCP servers cannot keep up with requests from legitimate DHCP clients. The address space of those servers is completely used up, so they can no longer assign IP addresses and lease times to clients. DHCP requests from those clients are either dropped—that is, the result is a denial of service (DoS)—or directed to a rogue DHCP server set up by the attacker to imitate a legitimate DHCP server.

    To protect the switch from DHCP starvation attacks, use the MAC limiting feature. Specify the maximum number of MAC addresses that the switch can learn on the access interfaces to which DHCP clients connect. The DHCP server or servers can then supply only the specified number of IP addresses over each of those interfaces. If a DHCP starvation attack occurs after the maximum number of IP addresses has been assigned, the attack fails.

    Modified: 2017-08-31