Extensible, asynchronous, centralized analysis and optimization of server responses to client requests

• US 9,906,620 B2
• Filed: 04/20/2009
• Issued: 02/27/2018
• Est. Priority Date: 05/05/2008
• Status: Active Grant

• Current Assignee: Radware Limited
• Original Assignee: Radware Limited
• Inventors: Alstad, Kent
• Primary Examiner(s): JAKOVAC, RYAN J
• Application Number:
  US12/426,909
• Publication Number:
  US 20090276488A1
• Time in Patent Office: 3,235 Days
• Field of Search: 709203
• CPC Class Codes: G06F 16/9574 of access to content, e.g. by caching
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: H04L 67/303 Terminal profiles
• CPC Class Codes: H04L 67/5651 Reducing the amount or size of exchanged application data

Abstract

An optimizer for messaging systems learns the purpose and context of each message and combines that information with knowledge of the specific client that will be rendering the response, such as a specific HTML browser. Any of a number of optimization factors can be applied, singly or in any combination. Messages are analyzed offline until a configurable threshold is reached, indicating that enough data has been sampled to develop a valid instruction set, to be applied to the responses that a server generates for a particular request. Responses are parsed into tokens and instructions for each type of token are compiled into instruction sets that are stored. These instructions sets continue to be iteratively improved as more data is collected, until the configurable sampling threshold is reached.

Claims

1. A computer-implemented method for optimizing a messaging system for communicating between at least one server and at least one client, comprising:

establishing a threshold representing an amount of data collection to generate at least one optimization instruction set;

intercepting a plurality of server-originated responses;

for each of the plurality of intercepted responses;

making a copy of the response; and

performing an analysis of the copy of the response to collect data for optimization by at least parsing the copy of the response to derive a plurality of tokens and classify the response-derived tokens into a plurality of predefined token categories;

responsive to the analysis and when having collected data to reach the established threshold,

generating at least one optimization instruction set, wherein each of the at least one optimization instruction set is based on a predefined token category from a plurality of predefined token categories having a particular token type, wherein each of the at least one optimization instruction set is generated for at least one content element common to responses associated with at least two users;

storing the generated at least one optimization instruction set in a storage device; and

intercepting at least one subsequent response originating at a server; and

applying the at least one generated optimization instruction set, based on a tokenization of the at least one subsequent response, to modify the at least one subsequent intercepted response so as to improve client performance when rendering the response;

wherein interception of any response is performed external to its originating server; and

wherein a timeout condition is established for expiry of at least one stored instruction set, and wherein, responsive to the timeout condition being met for a stored instruction set;

flagging the stored instruction set as expired;

initiating analysis of a plurality of copies of responses to collect data for optimization to replace the expired instruction set by at least parsing the copies of the responses to derive a new plurality of tokens and classifying the new responses derived tokens into a plurality of predefined token categories;

responsive to the analysis having collected data to reach the established threshold, generating a replacement optimization instruction set; and

storing the generated replacement optimization instruction set.

2. The method of claim 1, wherein the analysis of the copy of the response is performed concurrently with generating and transmitting responses.

3. The method of claim 1, wherein generating the at least one optimization instruction set comprises:

identifying a plurality of resources to be consolidated; and

generating instructions for consolidating the identified plurality of resources;

and wherein the optimization instruction set comprises instructions for replacing the plurality of resources with the consolidated resources.

4. The method of claim 3, wherein identifying a plurality of resources to be consolidated comprises at least one selected from the group consisting of:

identifying a plurality of cascading style sheet resources to be consolidated;

identifying a plurality of images to be consolidated; and

identifying a plurality of JavaScript resources to be consolidated.

5. The method of claim 1, wherein applying the at least one generated optimization instruction set comprises at least one selected from the group consisting of:

consolidating a plurality of cascading style sheets;

consolidating a plurality of images; and

consolidating a plurality of JavaScript resources.

6. The method of claim 1, wherein applying the at least one generated optimization instruction set comprises rewriting resource references to use a URI stem that increases the number of concurrent connections.

7. The method of claim 1, wherein applying the at least one generated optimization instruction set comprises aliasing references.

8. The method of claim 1, wherein applying the at least one generated optimization instruction set comprises:

predicting resource usage; and

preloading at least one resource based on the prediction.

9. The method of claim 1, wherein applying the at least one generated optimization instruction set comprises:

identifying static links within the subsequent response; and

transmitting the identified static links to the client prior to rendering of other portions of the subsequent response.

10. The method of claim 1, wherein applying the at least one generated optimization instruction set comprises:

identifying JavaScript code within the subsequent response; and

deferring at least one of loading and execution of the identified JavaScript code.

11. The method of claim 1, wherein each response comprises a server response to a request.

12. The method of claim 1, wherein each response comprises a response to an HTTP request.

13. The method of claim 1, wherein performing the analysis of the copy of the response and generating the at least one optimization instruction set are performed at an appliance located in a data path between the server and the client.

14. The method of claim 1, wherein:

performing the analysis of the copy of the response comprises generating the at least one optimization instruction set comprises compiling instructions for at least one token type.

15. The method of claim 1, wherein performing the analysis of the copy of the response comprises determining at least one portion of the copy of the response that is suitable for optimization.

16. The method of claim 1, further comprising:

performing additional analysis of a plurality of copies of responses to collect additional data for optimization; and

modifying at least one previously stored optimization instruction set based on the additional analysis.

47. The method of claim 1, wherein each predefined token category from a plurality of predefined token categories having a particular token type is based on a content-type of an intercepted response.

17. A system for optimizing a messaging system for communicating between at least one server and at least one client, comprising:

a computing device, for intercepting a plurality of server-originated responses and for making a copy of each intercepted response;

a compiler, for performing analysis of the copies of the intercepted responses to collect data for optimization, wherein the compiler is further configured to parse a copy of the response to derive a plurality of tokens and classify the response-derived tokens into predefined token categories, and responsive to the analysis and when having collected data to reach an established threshold, generate at least one optimization instruction set, wherein each of the at least one optimization instruction set is based on a predefined token category having a particular token type, wherein generating each of the at least one optimization instruction set is generated for at least one content element common to responses associated with at least two users;

a storage device, coupled to the compiler, for storing the at least one generated optimization instruction set; and

a rewriter, coupled to the storage device, for applying the at least one generated optimization instruction set to modify at least one subsequent response originating at a server and intercepted by the computing device, so as to improve client performance when rendering the response;

wherein applying the at least one generated optimization instruction is based on a tokenization of the at least one subsequent response; and

wherein interception of any response is performed external to its originating server; and

wherein a timeout condition is established for expiry of at least one stored instruction set, and wherein, responsive to the timeout condition being met for a stored instruction set;

the compiler flags the stored instruction set as expired;

the compiler initiates analysis of a plurality of copies of responses to collect data for optimization to replace the expired instruction set by at least parsing the copies of the responses to derive a new plurality of tokens and classifying the new responses derived tokens into a plurality of predefined token categories;

responsive to the analysis having collected data to reach the established threshold, the compiler generates a replacement optimization instruction set; and

the storage device stores the generated replacement optimization instruction set.

18. The system of claim 17, wherein the compiler generates the at least one optimization instruction set by:

identifying a plurality of resources to be consolidated; and

generating instructions for consolidating the identified plurality of resources;

and wherein the at least one optimization instruction set comprises instructions for

replacing the plurality of resources with the consolidated resources.

19. The system of claim 18, wherein the compiler identifies a plurality of resources to be consolidated by performing at least one selected from the group consisting of:

identifying a plurality of cascading style sheet resources to be consolidated;

identifying a plurality of images to be consolidated; and

identifying a plurality of JavaScript resources to be consolidated.

20. The system of claim 17, wherein the rewriter applies the at least one generated optimization instruction set by performing at least one selected from the group consisting of:

consolidating a plurality of cascading style sheets;

consolidating a plurality of images; and

consolidating a plurality of JavaScript resources.

21. The system of claim 17, wherein the rewriter applies the at least one generated optimization instruction set by rewriting resource references to use a URI stem that increases the number of concurrent connections.

22. The system of claim 17, wherein the rewriter applies the at least one generated optimization instruction set by aliasing references.

23. The system of claim 17, wherein the rewriter applies the at least one generated optimization instruction set by:

predicting resource usage; and

preloading at least one resource based on the prediction.

24. The system of claim 17, wherein the rewriter applies the at least one generated optimization instruction set by:

identifying static links within the subsequent response; and

transmitting identified static links to the client prior to rendering of other portions of the subsequent response.

25. The system of claim 17, wherein the rewriter applies the at least one generated optimization instruction set by:

identifying JavaScript code within the subsequent response; and

deferring at least one of loading and execution of identified JavaScript code.

26. The system of claim 17, wherein each response comprises a server response to a request.

27. The system of claim 17, wherein each response comprises a response to an HTTP request.

28. The system of claim 17, wherein the compiler is a component of an appliance located in a data path between the server and the client.

29. The system of claim 17, wherein:

the compiler performs analysis by parsing the copies of the responses to derive a plurality of response-derived tokens and classifying the response-derived tokens according to type; and

the compiler generates the at least one optimization instruction set by compiling instructions for at least one associated token type.

30. The system of claim 17, wherein the compiler performs analysis by, for at least one response, determining at least one portion of the copy of the response that is suitable for optimization.

31. The system of claim 17, wherein the compiler:

performs additional analysis of a plurality of copies of responses to collect additional data for optimization; and

modifies at least one previously stored the at least one optimization instruction set based on the additional analysis.

32. A computer program product for optimizing a messaging system for communicating between at least one server and at least one client, comprising:

a non-transitory computer-readable storage medium; and

computer program code, encoded on the medium, for;

establishing a threshold representing an amount of data collection to generate at least one optimization instruction set;

intercepting a plurality of server-originated responses

for each of the plurality of intercepted responses;

making a copy of the response; and

performing analysis of the copy of the response to collect data for optimization by at least parsing the copy of the response to derive a plurality of tokens and classifying the response-derived tokens into a plurality of predefined token categories;

responsive to the analysis and when having collected data to reach the established threshold, generating at least one optimization instruction set, wherein each of the at least one optimization instruction set is based on a predefined token category from a plurality of predefined token categories having a particular token type, wherein each of the at least one optimization instruction set is generated for at least one content element common to responses associated with at least two users;

storing the at least one generated optimization instruction set in a storage device;

intercepting at least one subsequent response originating at a server; and

applying the at least one generated optimization instruction set, based on a tokenization of the at least one subsequent response, to modify the at least one subsequent intercepted response so as to improve client performance when rendering the response

wherein interception of any response is performed external to its originating server; and

wherein a timeout condition is established for expiry of at least one stored instruction set, and wherein, responsive to the timeout condition being met for a stored instruction set;

flagging the stored instruction set as expired;

initiating analysis of a plurality of copies of responses to collect data for optimization to replace the expired instruction set by at least parsing the copies of the responses to derive a new plurality of tokens and classifying the new responses derived tokens into a plurality of predefined token categories;

responsive to the analysis having collected data to reach the established threshold, generating a replacement optimization instruction set; and

storing the generated replacement optimization instruction set.

33. The computer program product of claim 32, wherein the computer program code causes the analysis of the copy of the response to be performed concurrently with generating and transmitting responses.

34. The computer program product of claim 32, wherein the computer program code for generating the at least one optimization instruction set comprises computer program code for:

identifying a plurality of resources to be consolidated; and

generating instructions for consolidating the identified plurality of resources;

and wherein the at least one optimization instruction set comprises instructions for replacing the plurality of resources with the consolidated resources.

35. The computer program product of claim 34, wherein the computer program code for identifying a plurality of resources to be consolidated comprises computer program code for performing at least one selected from the group consisting of:

identifying a plurality of cascading style sheet resources to be consolidated;

identifying a plurality of images to be consolidated; and

identifying a plurality of JavaScript resources to be consolidated.

36. The computer program product of claim 32, wherein the computer program code for applying the generated at least one optimization instruction set comprises computer program code for performing at least one selecting from the group consisting of:

consolidating a plurality of cascading style sheets;

consolidating a plurality of images; and

consolidating a plurality of JavaScript resources.

37. The computer program product of claim 32, wherein the computer program code for applying the generated at least one optimization instruction set comprises computer program code for rewriting resource references to use a URI stem that increases the number of concurrent connections.

38. The computer program product of claim 32, wherein the computer program code for applying the generated at least one optimization instruction set comprises computer program code for aliasing references.

39. The computer program product of claim 32, wherein the computer program code for applying the generated at least one optimization instruction set comprises computer program code for:

predicting resource usage; and

preloading at least one resource based on the prediction.

40. The computer program product of claim 32, wherein the computer program code for applying the generated at least one optimization instruction set comprises computer program code for:

identifying static links within the subsequent response; and

transmitting the identified static links to the client prior to rendering of other portions of the subsequent response.

41. The computer program product of claim 32, wherein the computer program code for applying the generated at least one optimization instruction set comprises computer program code for:

identifying JavaScript code within the subsequent response; and

deferring at least one of loading and execution of identified JavaScript code.

42. The computer program product of claim 32, wherein each response comprises a server response to a request.

43. The computer program product of claim 32, wherein each response comprises a response to an HTTP request.

44. The computer program product of claim 32, wherein:

the computer program code for performing the analysis of the copy of the response comprises;

computer program code for parsing the copy of the response to derive the plurality of tokens and classifying the response-derived tokens into a plurality of predefined token categories; and

the computer program code for generating an optimization instruction set comprises;

computer program code for compiling instructions for at least one token type.

45. The computer program product of claim 32, wherein the computer program code for performing the analysis of the copy of the response comprises computer program code for determining at least one portion of the copy of the response that is suitable for optimization.

46. The computer program product of claim 32, further comprising computer program code for:

performing additional analysis of a plurality of copies of responses to collect additional data for optimization; and

modifying at least one previously stored optimization instruction set based on the additional analysis.