On-demand workload management in cloud bursting

Patent metadata

Field	Value
Publication number	US-10693963-B2
Application number	US-201816128910-A
Country	US
Kind code	B2
Filing date	Sep 12, 2018
Priority date	Oct 27, 2015
Publication date	Jun 23, 2020
Grant date	Jun 23, 2020

Abstract

A system, computer program product, and method is described to manage cloud bursting from a private cloud to a public cloud. The method starts with executing one or more software components of at least one application receiving transaction requests on at least one processing node accessing at least one database server. The one processing node is managed by at least one controller in a first deployment model of a computing infrastructure. The controller, the processing nodes and the database servers each include instrumentation for tracking of performance metrics thereof. A map is created of transaction requests received by the controller to data resources in the database by analyzing the set of common structures. The map is used to identify which of the software components and corresponding data resources in the database to move a processing node in a second deployment model of a computing infrastructure.

First claim

What is claimed is:

1. A computer-based method for on-demand workload management between a first deployment model of a computing infrastructure and a second deployment model of a computing infrastructure, the method comprising:
executing one or more software components of at least one application receiving transaction requests on a first processing node accessing at least one file server with persistent data in at least one database, wherein the transaction requests are tracked using tokens and each of the tokens include a thread identifier for a thread handling each of the transaction requests, the first processing node is managed by at least one controller in a first deployment model of a computing infrastructure;
performing enhanced transaction tracing analytics through use of instrumentation of the first processing node and the file server-for tracking which files in which table in which database in the first processing node in the file server are being requested;
parsing the transaction requests tracked by the instrumentation into a set of common structures for analysis;
creating a transaction-data map of transaction requests received by the controller to data resources in the file server by analyzing the set of common structures; and
using the transaction-data map to identify which of the one or more software components and corresponding data resources in the file server to move over to a second processing node in a second deployment model of a computing infrastructure.

2. The computer-based method of claim 1 , wherein the file server is at least one database server, each of the tokens include
a user identifier of a user.

3. The computer-based method of claim 1 , where in the first deployment model of a computing infrastructure is a private cloud, and
wherein the transaction-data map includes a timestamp, a method name, a table name in the at least one database, an inbound/outbound type with a source identifier,
wherein the using the transaction-data map to identify which of the one or more software components and corresponding data resources in the file server to move over to a second processing node in a second deployment model of a computing infrastructure includes using information from a load balancer and predefined weights assigned to the one or more software components.

4. The computer-based method of claim 3 , wherein the creating the transaction-data map includes creating the transaction-data map with a frequency of the inbound/outbound type with a source identifier during a time period, and a response time to fulfill the inbound/outbound type with a source identifier of transaction.

5. The computer-based method of claim 4 , wherein the creating the transaction-data map further includes creating the transaction-data map a settable weighting factor associated with each of the inbound/outbound type with a source identifier for each of the transaction requests.

6. The computer-based method of claim 5 , wherein using the transaction-data map includes identifying software components with lower frequency of the inbound/outbound type with a source identifier of each of the transaction requests during the time period.

7. A system for managing allocation of software components between a first deployment model of a computing infrastructure and a second deployment model of a computing infrastructure, the system comprising:
a memory;
a processor communicatively coupled to the memory, where the processor is configured to perform
executing one or more software components of at least one application receiving transaction requests on a first processing node accessing at least one file server with persistent data in at least one database, wherein the transaction requests are tracked using tokens and each of the tokens include a thread identifier for a thread handling each of the transaction requests, the first processing node is managed by at least one controller in a first deployment model of a computing infrastructure
performing enhanced transaction tracing analytics through use of instrumentation of the first processing node and the file server for tracking which files in which table in which database in the first processing node in the file server are being requested;
parsing the transaction requests tracked by the instrumentation into a set of common structures for analysis;
creating a transaction-data map of transaction requests received by the controller to data resources in the file server by analyzing the set of common structures; and
using the transaction-data map with information from a load balancer and predefined weights assigned to the one or more software components to identify which of the one or more software components and corresponding data resources in the file server to move over to a second processing node in a second deployment model of a computing infrastructure.

8. The system of claim 7 , wherein the file server is at least one database server, each of the tokens include
a user identifier of a user.

9. The system of claim 7 , where in the first deployment model of a computing infrastructure is a private cloud, and
wherein the transaction-data map includes a timestamp, a method name, a table name in the at least one database, an inbound/outbound type with a source identifier,
wherein the using the transaction-data map to identify which of the one or more software components and corresponding data resources in the file server to move over to a second processing node in a second deployment model of a computing infrastructure includes using information from a load balancer and predefined weights assigned to the one or more software components.

10. The system of claim 9 , the creating the transaction-data map includes creating the transaction-data map with a frequency of the inbound/outbound type with a source identifier during a time period, and a response time to fulfill the inbound/outbound type with a source identifier of transaction.

11. The system of claim 10 , wherein the creating the transaction-data map further includes creating the transaction-data map a settable weighting factor associated with each of the inbound/outbound type with a source identifier for each of the transaction requests.

12. The system of claim 11 , wherein using the transaction-data map includes identifying software components with lower frequency of the inbound/outbound type with a source identifier of transaction request during the time period.

13. A non-transitory computer program product for managing allocation of software components between a first deployment model of a computing infrastructure and a second deployment model of a computing infrastructure, the non-transitory computer program product comprising a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code configured to perform:
executing one or more software components of at least one application receiving transaction requests on a first processing node accessing at least one file server with persistent data in at least one database, wherein the transaction requests are tracked using tokens and each of the tokens include a thread identifier for a thread handling each of the transaction requests, the first processing node is managed by at least one controller in a first deployment model of a computing infrastructure;
performing enhanced transaction tracing analytics through use of instrumentation of the first processing node and the file server for tracking which files in which table in which database in the first processing node in the file server are being requested;
parsing the transaction requests tracked by

Assignees

• IBM

Inventors

• Liu Kai
• Sun Kewei
• Xu Di
• Zhong Xiao

Classifications

• H04L67/1097 Primary

  for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS] · CPC title

• H04L67/1029

  using data related to the state of servers by a load balancer · CPC title