The invention relates generally to the field of database searching and, more particularly, to a method and apparatus for enhanced hashing.
In networking systems, routers and/or switches typically move packets of information from one of a number of input ports to one or more output ports. Referring now to
A lookup function, such as may be implemented in Lookup Engine 104 or the like, can include a content addressable memory (CAM), but this approach may be relatively expensive. Another approach is to use a standard memory, such as static random-access memory (SRAM), with a direct indexing of the entries. However, because a typical key may be, for example, 60-bits wide, a 260 universe (i.e., 260 table entries) would be required and this approach may be impractical and/or cost-prohibitive. A more common approach is to use the standard memory as a hashing table, which essentially provides a “many-to-one” function so that the table size can be made substantially smaller and the overall system cost can be reduced.
Referring now to
Consequently, what is needed is an enhanced hashing approach that does not include an ECC field in the hash table, but still maintains an ECC function.
The invention overcomes the identified limitations and provides a hashing solution with efficient table entry storage and other advantageous features.
According to embodiments of the invention, a search key lookup system can include a hash table having a plurality of entries and a function generator. The function generator can be coupled to the table and configured to receive a key and to provide a first function and a second function. The first function can be a Cyclic Redundancy Code (CRC) type function and the second function can be an Error Checking and Correcting (ECC) type function. Further, an address of the table can include a concatenation of the results of the CRC and the ECC type functions.
According to another aspect of embodiments of the invention, a method of searching a hash table can include the steps of providing a first function of an incoming key, providing a second function of the incoming key, extracting a stored key from the table to provide an extracted key, performing a correction on the extracted key to provide a corrected key, and comparing the corrected key with the incoming key. The first function can be an ECC type function and the second function can be a CRC type function. Further, the correction can include an ECC type correction.
Advantages of the invention include providing efficient table entry storage while maintaining ECC functionality.
Embodiments of the invention are described with reference to the FIGS, in which:
Embodiments of the invention are described with reference to specific diagrams depicting system arrangements and methods. Those skilled in the art will recognize that the description is for illustration and to provide the best mode of practicing the invention. The description is not meant to be limiting. For example, reference is made to specific function generators, such as Cyclic Redundancy Code (CRC) and Error Checking and Correcting (ECC) type functions, but the invention is applicable to other types of functions and/or mappings as well. Also, hash tables and the specific fields and/or arrangements thereof in a system merely provide example implementations and should not be construed as limiting. Further, the standard of corruption protection used in examples of this approach is single-bit-error correction and double-bit-error detection (SECDED), but the invention is applicable to other standards and/or types of corruption protection as well.
Referring now to
Referring now to
As an example application, a Media Access Control (MAC) address may be 48-bits wide, but the system application may only require storage of a subset of the possible keys. Thus, a set of “actually stored” keys may be substantially less than the 248 “universe” of possible keys. An attribute, such as stored in the field Attribute 358, for each actually stored key can include a port number associated with the corresponding MAC address, for example. Further, this approach is applicable to common collision resolution methods, as used in typical hash table implementations.
Key insertion and search procedures can be described with reference to pseudo-code. Pseudo-code for inserting a key into the hash table can include:
Pseudo-code for searching for a key in a hash table can include:
Referring now to
The searching of the hash table for a key procedure can include the flow proceeding from step Generate CRC 408 to step Extract Key 412. The key extraction can include reading the memory at an address location referenced by the address formed by the concatenation of CRC and ECC function outputs, for example. Next, the flow can proceed to step ECC Correct Key 414. This ECC correction can provide an ECC corrected version of the extracted key. Next, the flow can proceed to step Compare Incoming Key To ECC Correct Key 416. This step can determine if a match or a mismatch occurred as a result of the search. Accordingly, if the incoming key was not found in the hash table, a mismatch indication can be provided. However, if the incoming key was found in the hash table, a match indication can be provided. The searching procedure can end in Done 418. Because many keys may be mapped into the same entry of the table, key information can be stored for each entry. Accordingly, this comparison with the stored key itself, or an ECC corrected version thereof, may be needed in order to determine if there is a true match in the hash table.
Advantages of the invention include providing efficient table entry storage while maintaining ECC functionality.
Having disclosed exemplary embodiments and the best mode, modifications and variations may be made to the disclosed embodiments while remaining within the subject and spirit of the invention as defined by the following claims.
The present application is a continuation of an application filed Nov. 7, 2003 now U.S. Pat. No. 7,617,241 under Ser. No. 10/703,842, which is incorporated herein by reference.
Number | Name | Date | Kind |
---|---|---|---|
5491700 | Wright et al. | Feb 1996 | A |
6467060 | Malakapalli et al. | Oct 2002 | B1 |
6651099 | Dietz et al. | Nov 2003 | B1 |
6738935 | Kimmitt | May 2004 | B1 |
6904057 | Sarkinen et al. | Jun 2005 | B2 |
7162678 | Saliba | Jan 2007 | B2 |
7234019 | Kao et al. | Jun 2007 | B1 |
7617241 | Yang | Nov 2009 | B2 |
20030081615 | Kohn et al. | May 2003 | A1 |
20030097520 | Lai et al. | May 2003 | A1 |
Entry |
---|
Cormen et al., “Introduction to Algorithms” The MIT Press, Cambridge, Massachusetts London, England, 2001. |
Non-Final Office Action dated Jul. 21, 2009 for U.S. Appl. No. 11/801,937. |
Number | Date | Country | |
---|---|---|---|
20100017681 A1 | Jan 2010 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 10703842 | Nov 2003 | US |
Child | 12567608 | US |