CLOUD MESSAGING SERVICES OPTIMIZATION THROUGH ADAPTIVE MESSAGE COMPRESSION

• US 20140337442A1
• Filed: 05/10/2013
• Published: 11/13/2014
• Est. Priority Date: 05/10/2013
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Chen, Yao-Min, Zhuang, Zhenyun
• Granted Patent: US 9,270,624 B2
• Time in Patent Office: Days
• US Class Current: 709/206
• CPC Class Codes: H04L 51/06 Message adaptation to terminal or network requirements
• CPC Class Codes: H04L 51/066 Format adaptation, e.g. format conversion or compression
• CPC Class Codes: H04L 69/04 Protocols for data compression, e.g. ROHC

Abstract

Intelligent adjustment of compression levels of messages being brokered or otherwise handled by message service providers in a cloud-computing environment to strike a balance between the reduction in network usage and storage requirement resources owing to message compression on one hand and the increase in CPU resource usage owing to message compression on the other hand for reducing usage-based charges from the cloud customer'"'"'s point of view and reducing the equipment and operational cost from the cloud provider'"'"'s point of view (resulting in the ability of the provider to take on and/or retain customers). Determination of an “optimal” compression level that strikes an appropriate balance among resource usage may include consideration of a current messaging load (e.g., bytes of data per unit time), a current compression level, and current usage levels of one or more of the various resources.

Claims

1. A method, comprising:

obtaining, over time, by a messaging service provider that facilitates the transmission of messages between message producers and message consumers, respective data collections representing a plurality of messaging states, wherein each messaging state data collection includes a messaging load of messages being received at the message service provider, a level of compression of the received messages by the messaging service provider at the messaging load, and one or more usage levels of a plurality of resources being consumed as a function of the messaging load and messaging compression level;

estimating, from the collected messaging state data collections, resource usage levels of the plurality of resources at a plurality of different messaging compression levels based on a current messaging load;

selecting a particular messaging compression level from the plurality of different messaging compression levels corresponding to desired resource usage levels; and

compressing copies of messages received at the messaging service provider under the current messaging load at the particular messaging compression level.

2. The method of claim 1, wherein each resource of the plurality of resources is associated with a respective cost for usage of the resource, and wherein the selecting step comprises:

identifying a particular messaging compression level from the plurality of different messaging compression levels that results in a desired overall cost for use of the plurality of resources.

3. The method of claim 2, wherein the plurality of resources includes at least two selected from the group consisting of processing, storage, and networking.

4. The method of claim 3, wherein the resource usage levels of the storage and networking are collectively considered during the obtaining step, estimating step and selecting step.

5. The method of claim 1, wherein the obtaining step comprises:

storing each messaging state data collection in a respective record of a vector, wherein the estimating step utilizes at least some of the records.

6. The method of claim 5, wherein the at least some of the records are more recent to a current time than are other of the records.

7. The method of claim 1, wherein the collected messaging state data collections include a plurality of data points, and wherein the estimating step comprises:

performing a linear regression on the plurality of data points to predict resource usage levels of the plurality of resources at the plurality of different messaging compression levels based on the current messaging load.

8. The method of claim 7, wherein data points of the plurality of data points associated with messaging loads and compression levels that are closer to a particular messaging load and compression level under consideration in the estimating than are messaging loads and compression levels of other data points of the plurality of data points are weighted more heavily in the linear regression than are the other data points of the plurality of data points.

9. The method of claim 1, further comprising:

queuing the messages in a memory of the messaging service provider before the compressing step.

10. The method of claim 9, further comprising:

sending the compressed copies of the queued messages to a storage server over at least one network after the compressing step.

11. The method of claim 1, wherein each compression level comprises a percentage reduction in size of each received message copy.

12. The method of claim 1, wherein each compression level comprises a percentage of all received message copies to be compressed.

13. An adaptive message compression system for a message service provider in a cloud-computing environment, comprising:

a queue that stores messages received from a message producer over at least one network for retrieval by a message consumer over at least one network;

a compression engine that compresses copies of the queued messages at one or more compression levels, the compressed message copies being transmitted to a storage server over at least one network for storage, wherein compression and transmission of the of the queued message copies to the storage server consumes processing, network bandwidth and storage resources of the adaptive message compression system; and

an analysis module that determines at least one particular compression level that reduces collective costs associated with consumption of the processing, network bandwidth and storage resources of the adaptive message compression system, wherein the particular compression level is implemented by the compression module.

14. The system of claim 13, wherein the analysis module includes:

a collection module that records a messaging load, a messaging compression level, and usage levels of the processing, network bandwidth and storage resources after each of a plurality of different time periods;

an estimation module that predicts usage levels of the processing, network bandwidth and storage resources for each of a plurality of different compression levels based on a current messaging load; and

a selection module that identifies the particular compression level from the plurality of different compression levels.

15. The system of claim 14, wherein the estimation module incorporates a least-squares fitting to solve a linear regression model.

16. The system of claim 13, further comprising:

a triggering module that detects when a message has been retrieved from the queue by a message consumer and, in response, triggers the storage server to delete the compressed copy of the message stored in the storage server.

17. The system of claim 13, wherein each compression level comprises a percentage reduction in size of each received message and/or a percentage of all received messages to be compressed.

18. The system of claim 13, wherein the compression engine compresses the copies of the queued messages at the one or more compression levels using one of a plurality of compression techniques, and wherein the analysis module selects the one of the compression techniques so as to reduce collective costs associated with consumption of the processing, network bandwidth and storage resources of the adaptive message compression system.

19. A cloud-based computing system, comprising:

a processing module; and

a memory module logically connected to the processing module and comprising a set of computer readable instructions executable by the processing module to;

receive messages from message producers over one or more networks at one or more messaging loads;

store the messages in the memory module, the messages intended for retrieval by message consumers over one or more networks;

determine at least one particular compression level at which to compress copies of the messages to be sent for storage on a network-based storage server, the particular compression level being a compression level that reduces overall costs associated with consumption of processing, network bandwidth and storage resources of the system in conjunction with compression and storage of messages;

compress copies of the stored messages at the particular compression level; and

send the compressed message copies to a network-based storage server for storage.

20. The system of claim 19, wherein the computer readable instructions are executable by the processor to determine by way of:

recording a messaging load, a messaging compression level, and usage levels of the processing, network bandwidth and storage resources after each of a plurality of different time periods;

predicting usage levels of the processing, network bandwidth and storage resources for each of a plurality of different compression levels based on a current messaging load; and

identifying the particular compression level from the plurality of different compression levels.