Hardware filtering support for denial-of-service attacks

• US 7,411,957 B2
• Filed: 03/26/2004
• Issued: 08/12/2008
• Est. Priority Date: 03/26/2004
• Status: Expired due to Fees

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Garner, Trevor, Stacy, John Kenneth, Hughes, Martin W., Lee, William R.
• Primary Examiner(s): LEE, CHI HO A
• Application Number:
  US10/811,195
• Publication Number:
  US 20050213570A1
• Time in Patent Office: 1,600 Days
• Field of Search: None
• US Class Current: 370/392
• CPC Class Codes: H04L 47/32 by discarding or delaying data units, e.g. packets or frames
• CPC Class Codes: H04L 49/90 Buffering arrangements
• CPC Class Codes: H04L 49/901 using storage descriptor, e.g. read or write pointers
• CPC Class Codes: H04L 49/9042 Separate storage for different parts of the packet, e.g. header and payload
• CPC Class Codes: H04L 49/9047 including multiple buffers, e.g. buffer pools
• CPC Class Codes: H04L 63/1416 Event detection, e.g. attack signature detection
• CPC Class Codes: H04L 63/1458 Denial of Service

Abstract

A system and method is provided for automatically identifying and removing malicious data packets, such as denial-of-service (DoS) packets, in an intermediate network node before the packets can be forwarded to a central processing unit (CPU) in the node. The CPU'"'"'s processing bandwidth is therefore not consumed identifying and removing the malicious packets from the system memory. As such, processing of the malicious packets is essentially “off-loaded” from the CPU, thereby enabling the CPU to process non-malicious packets in a more efficient manner. Unlike prior implementations, the invention identifies malicious packets having complex encapsulations that can not be identified using traditional techniques, such as ternary content addressable memories (TCAM) or lookup tables.

Claims

1. A method for a network node, which includes a central processing unit (CPU) configured to execute a router operating system, to filter malicious data packets received at the network node, the method comprising:

receiving a data packet at the network node;

performing hash-based flow classification on the received data packet to determine whether the received data packet is a malicious data packet; and

discarding, by a hardware assist (HWA) module of a system controller that is coupled to the CPU, the received data packet before the data packet can be forwarded to the CPU for processing by the router operating system, if the received data packet is determined to be a malicious data packet.

2. The method of claim 1, wherein the step of performing hash-based flow classification further comprises:

identifying a packet type associated with the received data packet;

extracting a set of signature information corresponding to the identified packet type; and

searching a hash table to locate the extracted set of signature information.

3. The method of claim 2, further comprising:

configuring the hash table, either manually or automatically, to associate the set of signature information with a data flow; and

determining whether the data flow associated with the set of signature information corresponds to a malicious data flow.

4. The method of claim 1, further comprising:

associating the received data packet with a destination in the network node as a result of the hash-based flow classification.

5. The method of claim 4, further comprising:

determining whether the destination associated with the received data packet is a predetermined destination associated with malicious data packets.

6. The method of claim 5, further comprising:

in response to determining that the destination associated with the received data packet is the predetermined destination, performing the steps of;

removing buffer pointers from a set of descriptors associated with the received data packet; and

storing the removed buffer pointers on a queue of free buffer pointers.

7. The method of claim 6, further comprising:

if the queue of free buffer pointers does not contain enough available entries to store the removed buffer pointers, storing the set of descriptors associated with the received data packet on a delete queue until enough entries become available in the queue of free buffer pointers.

8. The method of claim 6, further comprising:

transferring free buffer pointers from the router operating system to the queue of free buffer pointers.

9. The method of claim 1, wherein the step of performing hash-based flow classification is used in conjunction with an access control list or an intrusion detection system.

10. The method of claim 1, wherein the network node is an intermediate network node.

11. A network node, comprising:

a central processing unit (CPU) configured to execute instructions that implement a router operating system;

a network interface adapted to receive a data packet;

a memory having a plurality of storage locations addressable by the CPU, the storage locations being configured to store;

(i) at least a portion of the router operating system instructions,

(ii) one or more data buffers for storing the received data packet, and

(iii) a searchable data structure configured to store information associated with the received data packet; and

a system controller coupled to the memory and the CPU, the system controller including a hardware assist (HWA) module configured to discard malicious data packets from the network node before the malicious data packets can be forwarded to the CPU for processing by the router operating system.

12. The network node of claim 11, wherein the searchable data structure is a hash table.

13. The network node of claim 11, wherein the HWA module includes a direct memory access (DMA) controller and a flow classifier.

14. The network node of claim 13, wherein the DMA controller includes:

an ingress descriptor first in, first out (FIFO) queue configured to store a set of descriptors referencing the one or more data buffers in which the received data packet is stored;

a packet-header buffer configured to store information contained in at least one packet header prepended to the received data packet;

an egress descriptor FIFO configured to store the set of descriptors as well as a data flow identification (ID) value for identifying the data flow associated with the received data packet and a destination value for identifying a destination in the network node associated with the received data packet, the flow classifier searching the searchable data structure to locate the data flow ID value and the destination value; and

a free-buffer FIFO containing a set of free buffer descriptors allocated for the network interface.

15. The network node of claim 13, wherein the flow classifier includes:

a packet-identifier engine configured to identify a packet type associated with the received data packet based on information received from the DMA controller;

a signature-extraction engine configured to extract a set of signature information from a predetermined set of fields in the information received from the DMA controller, the predetermined set of fields being selected based on the packet type identified by the packet-identifier engine;

an address generator configured to generate a memory address based on the set of signature information, the memory address corresponding to an entry in the searchable data structure; and

a search module configured to search the searchable data structure to locate a flow ID value and a destination value associated with the received data packet.

16. The network node of claim 15, wherein the flow classifier further includes

an egress packet manager configured to reformat descriptors from an ingress descriptor format to an egress descriptor format.

17. The network node of claim 11, wherein the network node is an intermediate network node.

18. A network node including a central processing unit (CPU) configured to execute a router operating system, the network node comprising:

means for receiving a data packet at the network node;

means for performing hash-based flow classification on the received data packet to determine whether the received data packet is a malicious data packet; and

means for discarding, at a hardware assist (HWA) module of a system controller that is coupled to the CPU, the received data packet before the data packet can be forwarded to the CPU for processing by the router operating system, if the received data packet is determined to be a malicious data packet.

19. A computer-readable media including instructions for execution by a processor, the instructions for a method of filtering malicious data packets received at a network node in which a central processing unit (CPU) is configured to execute a router operating system, the method comprising:

receiving a data packet at the network node;

performing hash-based flow classification on the received data packet to determine whether the received data packet is a malicious data packet; and

discarding, by a hardware assist (HWA) module of a system controller that is coupled to the CPU, the received data packet before the data packet can be forwarded to the CPU for processing by the router operating system, if the received data packet is determined to be a malicious data packet.

20. A method comprising:

receiving a data packet;

storing the data packet in one or more data buffers referenced by a set of one or more packet descriptors;

passing the set of packet descriptors to an ingress descriptor queue at a direct memory access (DMA) controller;

performing, by a flow classifier coupled of the DMA controller, hash-based flow classification on the data packet referenced by the set of packet descriptors to associate the data packet with a particular data flow identification (ID) value;

passing the set of packet descriptors to an egress descriptor queue; and

determining based on the data flow ID value if the data packet is a malicious data packet and, if so, freeing the set of packet descriptors to discard the data packet before the data packet can be processed by a router operating system operating on a CPU coupled to the DMA controller and, if not, allowing the data packet to be processed by the router is operating system operating on the CPU coupled to the DMA controller.

21. The method of claim 20 wherein freeing the set of packet descriptors comprises:

removing buffer pointers from the set of packet descriptors and placing the buffer pointers on a free-buffer queue.

22. The method of claim 21, further comprising:

if the free-buffer queue has insufficient space to store additional removed buffer pointers, storing the set of descriptors in a delete queue until space becomes available in the free-buffer queue.

23. The method of claim 20 wherein allowing the data packet to be processed comprises:

sending an interrupt to the CPU to notify the CPU that the set of packet descriptors are available in the egress queue.

24. The method of claim 20, wherein performing hash-based flow classification comprises:

identifying a packet type associated with the data packet;

extracting a set of signature information corresponding to the identified packet type; and

searching a hash table to locate the extracted set of signature information and accessing one or more associated entries.

25. The method of claim 20 further comprising:

reformatting the set of packet descriptors from an ingress descriptor format to an egress descriptor format before passing the set of packet descriptors to the egress descriptor queue.