Dynamic content caching

• US 9,400,851 B2
• Filed: 06/23/2011
• Issued: 07/26/2016
• Est. Priority Date: 06/23/2011
• Status: Active Grant

• Current Assignee: Imperva Incorporated (Thales SA)
• Original Assignee: Incapsula Incorporated (Thales SA)
• Inventors: Zingerman, Boris, Almagor, Ohad, Shatz, Gur
• Primary Examiner(s): Birkhimer, Christopher D
• Application Number:
  US13/167,130
• Publication Number:
  US 20120331228A1
• Time in Patent Office: 1,860 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 12/0802 Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
• CPC Class Codes: G06F 12/0862 with prefetch
• CPC Class Codes: G06F 12/0864 using pseudo-associative means, e.g. set-associative or hashing
• CPC Class Codes: G06F 12/0866 for peripheral storage systems, e.g. disk cache
• CPC Class Codes: G06F 12/0897 with two or more cache hierarchy levels with multilevel cache hierarchies G06F12/0811
• CPC Class Codes: G06F 16/9574 of access to content, e.g. by caching

Abstract

A system for caching content including a server supplying at least one of static and non-static content elements, content distinguishing functionality operative to categorize elements of the non-static content as being either dynamic content elements or pseudodynamic content elements, and caching functionality operative to cache the pseudodynamic content elements. The static content elements are content elements which are identified by at least one of the server and metadata associated with the content elements as being expected not to change, the non-static content elements are content elements which are not identified by the server and/or by metadata associated with the content elements as being static content elements, the pseudodynamic content elements are non-static content elements which, based on observation, are not expected to change, and the dynamic content elements are non-static content elements which are not pseudodynamic.

Claims

1. A system for caching content comprising:

a server supplying at least one of static and non-static content elements;

content distinguishing functionality operative to categorize elements of said non-static content as being either dynamic content elements or pseudodynamic content elements; and

caching functionality operative to cache said pseudodynamic content elements;

said static content elements being content elements which are identified by at least one of said server and metadata associated with said content elements as being expected not to change for at least a predetermined period of time;

said non-static content elements being content elements which are not identified by said server and/or by metadata associated with said content elements as being static content elements;

said pseudodynamic content elements being non-static content elements which, based on observation of a digest value of the byte sequence thereof, are not expected to change, said observation of said digest value of said byte sequence comprising a combination of;

comparing a stored digest value of a byte sequence of at least one historical response to at least one historical request for a URL with a current digest value of a byte sequence of a current response to a current request for said URL;

comparing a timestamp of said stored digest value of said byte sequence of said at least one historical response to said at least one historical request for said URL with a timestamp of said current digest value of said byte sequence of said current response to said current request for said URL; and

ascertaining that a predefined sufficient variety of client attributes have been stored in association with said at least one historical request for said URL, said sufficient variety of client attributes comprising at least a predefined number of distinct IP addresses of clients originating said at least one historical request; and

said dynamic content elements being non-static content elements which are not pseudodynamic.

2. A system for caching content according to claim 1 and wherein said content distinguishing functionality is operative to distinguish between said dynamic content and said pseudodynamic content by distinguishing between content downloaded to disparate clients that has a changing byte content and content that has a static byte content.

3. A system for caching content according to claim 1 and wherein said caching functionality is operative to ascertain whether said content is cached upon receiving a content request from a client, said content request comprising a full URL.

4. A system for caching content according to claim 3 and wherein said caching functionality is operative to provide said content to said client upon ascertaining that said content is cached.

5. A system for caching content according to claim 3 and wherein said caching functionality is operative to route said content request to a web server which hosts said full URL upon ascertaining that said content is not cached.

6. A system for caching content according to claim 5 and wherein said client is operative to receive a response from said web server to said content request.

7. A system for caching content according to claim 5 and wherein said content distinguishing functionality is operative to receive a response from said web server to said content request.

8. A system for caching content according to claim 7 and wherein said content distinguishing functionality comprises a URL database which stores a list of previously processed full URLs which were previously processed by said content distinguishing functionality.

9. A system for caching content according to claim 8 and wherein said URL database stores a caching state associated with said previously processed full URL for each previously processed full URL stored in said URL database.

10. A system for caching content according to claim 9 and wherein said content distinguishing functionality is also operative to receive said full URL and attributes of said client.

11. A system for caching content according to claim 10 and wherein said attributes include source IP, session ID, user agent and screen size.

12. A system for caching content according to claim 9 and wherein said content distinguishing functionality is also operative to receive a digest of said response and a timestamp of said response.

13. A system for caching content according to claim 9 and wherein said content distinguishing functionality is also operative to calculate a digest of said response and a timestamp of said response.

14. A system for caching content according to claim 9 and wherein said content distinguishing functionality is also operative to ascertain whether said full URL is stored in said URL database.

15. A system for caching content according to claim 14 and wherein said content distinguishing functionality is operative to store said full URL, said digest, said timestamp and said attributes in said URL database and to set a stored caching state corresponding to said full URL to “

learning”

when said full URL is not stored in said URL database.

16. A system for caching content according to claim 15 and wherein said content distinguishing functionality is also operative to ascertain whether said stored caching state corresponding to said full URL is one of “

learning”

, “

pseudodynamic” and

“

dynamic”

when said full URL is stored in said URL database.

17. A system for caching content according to claim 16 and wherein said content distinguishing functionality is also operative to ascertain whether said digest is identical to a stored digest corresponding to said URL when said stored caching state is “

learning”

.

18. A system for caching content according to claim 17 and wherein said content distinguishing functionality is also operative to store said attributes in said URL database when said digest is identical to a stored digest corresponding to said URL, and responsive to a predefined learning time having elapsed since said stored caching state was initially set as “

learning” and

a predefined sufficient variety of client attributes having been stored over a predefined minimum number of responses associated with said URL said content distinguishing functionality is also operative to set said caching state as “

pseudodynamic”

, and to store said response in said cache with a predefined caching expiration time.

19. A system for caching content according to claim 17 and wherein said caching state of said URL is set as “

dynamic” and

said timestamp is stored when said digest is not identical to a stored digest corresponding to said URL.

20. A system for caching content according to claim 18 and wherein said variety of client attributes comprises a minimum number of distinct IP addresses in combination with a minimum number of distinct user agents.

21. A system for caching content according to claim 18 and wherein said predefined caching expiration time is shorter than said learning time.

22. A system for caching content according to claim 16 and wherein said content distinguishing functionality is also operative to ascertain whether said digest is identical to a stored digest corresponding to said URL when said stored caching state is “

pseudodynamic”

.

23. A system for caching content according to claim 22 and wherein said caching state of said URL is set as “

learning” and

said digest, said timestamp and said attributes are stored when said digest is not identical to a stored digest corresponding to said URL.

24. A system for caching content according to claim 22 and wherein said response is stored in said cache with a new caching expiration time associated therewith when said digest is identical to a stored digest corresponding to said URL.

25. A system for caching content according to claim 24 and wherein said new caching expiration time is equal to said predefined caching expiration time.

26. A system for caching content according to claim 24 and wherein said new caching expiration time is not equal to said predefined caching expiration time.

27. A system for caching content according to claim 16 and wherein said content distinguishing functionality is also operative to ascertain whether a predefined amount of refresh time has elapsed since a stored timestamp of said URL when said stored caching state is “

dynamic”

.

28. A system for caching content according to claim 27 and wherein said caching state of said URL is set as “

learning” and

said digest, said timestamp and said attributes are stored in the entry when a predefined amount of refresh time has elapsed since a stored timestamp of said URL.

29. A method for caching content comprising:

supplying at least one of static and non-static content elements;

categorizing elements of said non-static content as being either dynamic content elements or pseudodynamic content elements; and

caching said pseudodynamic content;

said static content elements being content elements which are identified by at least one of said server and metadata associated with said content elements as being expected not to change for at least a predetermined period of time;

said non-static content elements being content elements which are not identified by said server and/or by metadata associated with said content elements as being static content elements;

said pseudodynamic content elements being non-static content elements which, based on observation of a digest value of the byte sequence thereof, are not expected to change, said observation of said digest value of said byte sequence comprising a combination of;

comparing a stored digest value of a byte sequence of at least one historical response to at least one historical request for a URL with a current digest value of a byte sequence of a current response to a current request for said URL;

comparing a timestamp of said stored digest value of said byte sequence of said at least one historical response to said at least one historical request for said URL with a timestamp of said current digest value of said byte sequence of said current response to said current request for said URL; and

ascertaining that a predefined sufficient variety of client attributes have been stored in association with said at least one historical request for said URL, said sufficient variety of client attributes comprising at least a predefined number of distinct IP addresses of clients originating said at least one historical request; and

said dynamic content elements being non-static content elements which are not pseudodynamic.

30. A method for caching content according to claim 29 and wherein said categorizing comprises distinguishing between said dynamic content and said pseudodynamic content by distinguishing between content downloaded to disparate clients that has a changing byte content and content that has a static byte content.

31. A method for caching content according to claim 29 and wherein said categorizing also comprises ascertaining whether said content is cached upon receiving a content request comprising a full URL from a client.

32. A method for caching content according to claim 31 and wherein said categorizing comprises providing said content to said client upon ascertaining that said content is cached.

33. A method for caching content according to claim 31 and wherein said categorizing comprises routing said content request to a web server which hosts said full URL upon ascertaining that said content is not cached.

34. A method for caching content according to claim 33 and wherein said client is operative to receive a response from said web server to said content request.

35. A method for caching content according to claim 33 and wherein said categorizing also comprises receiving a response from said web server to said content request.

36. A method for caching content according to claim 35 and wherein said categorizing also comprises storing a list of previously processed full URLs.

37. A method for caching content according to claim 36 and wherein a caching state associated with said previously processed full URL is stored for each stored previously processed full URL.

38. A method for caching content according to claim 37 and wherein said categorizing also comprises receiving said full URL and attributes of said client.

39. A method for caching content according to claim 38 and wherein said attributes include source IP, session ID, user agent and screen size.

40. A method for caching content according to claim 37 and wherein said categorizing also comprises receiving a digest of said response and a timestamp of said response.

41. A method for caching content according to claim 37 and wherein said categorizing also comprises calculating a digest of said response and a timestamp of said response.

42. A method for caching content according to claim 37 and wherein said categorizing also comprises ascertaining whether said full URL is stored in said URL database.

43. A method for caching content according to claim 42 and wherein said categorizing comprises storing said full URL, said digest, said timestamp and said attributes and setting a stored caching state corresponding to said full URL to “

learning”

when said full URL is not stored.

44. A method for caching content according to claim 43 and wherein said categorizing also comprises ascertaining whether said stored caching state corresponding to said full URL is one of “

learning”

, “

pseudodynamic” and

“

dynamic”

when said full URL is stored.

45. A method for caching content according to claim 44 and wherein said categorizing also comprises ascertaining whether said digest is identical to a stored digest corresponding to said URL when said stored caching state is “

learning”

.

46. A method for caching content according to claim 45 and wherein said caching state of said URL is set as “

dynamic” and

said timestamp is stored when said digest is not identical to a stored digest corresponding to said URL.

47. A method for caching content according to claim 45 and wherein said categorizing also comprises storing said attributes when said digest is identical to a stored digest corresponding to said URL, and responsive to a predefined learning time having elapsed since said stored caching state was initially set as “

learning” and

a predefined sufficient variety of client attributes having been stored over a predefined minimum number of responses associated with said URL said categorizing also comprises setting said caching state as “

pseudodynamic”

, and caching said response with a predefined caching expiration time.

48. A method for caching content according to claim 47 and wherein said predefined caching expiration time is shorter than said learning time.

49. A method for caching content according to claim 47 and wherein said variety of client attributes comprises a minimum number of distinct IP addresses in combination with a minimum number of distinct user agents.

50. A method for caching content according to claim 44 and wherein said categorizing also comprises ascertaining whether said digest is identical to a stored digest corresponding to said URL when said stored caching state is “

pseudodynamic”

.

51. A method for caching content according to claim 50 and wherein said caching state of said URL is set as “

learning”

, and said digest, said timestamp and said attributes are stored when said digest is not identical to a stored digest corresponding to said URL.

52. A method for caching content according to claim 50 and wherein said response is cached with a new caching expiration time associated therewith when said digest is identical to a stored digest corresponding to said URL.

53. A method for caching content according to claim 52 and wherein said new caching expiration time is not equal to said predefined caching expiration time.

54. A method for caching content according to claim 52 and wherein said new caching expiration time is equal to said predefined caching expiration time.

55. A method for caching content according to claim 44 and wherein, said categorizing also comprises ascertaining whether a predefined amount of refresh time has elapsed since a stored timestamp of said URL when said stored caching state is “

dynamic”

.

56. A method for caching content according to claim 55 and wherein said caching state of said URL is set as “

learning”

, and said digest, said timestamp and said attributes are stored in the entry when a predefined amount of refresh time has elapsed since a stored timestamp of said URL.

57. A content distinguishing method operative in a system for serving content comprising:

a server supplying at least one of static content and non-static content, said content distinguishing method comprising categorizing elements of said non-static content as being either dynamic content elements or pseudodynamic content elements; and

caching functionality operative to cache said pseudodynamic content;

said static content comprising content elements which are identified by at least one of said server and metadata associated with said content elements as being expected not to change for at least a predetermined period of time;

said non-static content comprising content elements which are not identified by said server and/or by metadata associated with said content elements as being static content elements;

said pseudodynamic content comprising non-static content elements which, based on observation of a digest value of the byte sequence thereof, are not expected to change, said observation of said digest value of said byte sequence comprising a combination of;

comparing a stored digest value of a byte sequence of at least one historical response to at least one historical request for a URL with a current digest value of a byte sequence of a current response to a current request for said URL;

comparing a timestamp of said stored digest value of said byte sequence of said at least one historical response to said at least one historical request for said URL with a timestamp of said current digest value of said byte sequence of said current response to said current request for said URL; and

ascertaining that a predefined sufficient variety of client attributes have been stored in association with said at least one historical request for said URL, said sufficient variety of client attributes comprising at least a predefined number of distinct IP addresses of clients originating said at least one historical request; and

said dynamic content comprising non-static content elements which are not pseudodynamic.

58. A content distinguishing method according to claim 57 and wherein said categorizing comprises distinguishing between said dynamic content and said pseudodynamic content by distinguishing between content downloaded to disparate clients that has a changing byte content and content that has a static byte content.