Methods and systems for batch processing in an on-demand service environment

• US 8,584,124 B2
• Filed: 03/31/2011
• Issued: 11/12/2013
• Est. Priority Date: 04/20/2010
• Status: Active Grant

• Current Assignee: Salesforce.com, Inc.
• Original Assignee: Salesforce.com, Inc.
• Inventors: Fee, Gregory D., Gallagher, William J
• Primary Examiner(s): Tang, Kenneth
• Application Number:
  US13/076,794
• Publication Number:
  US 20110258630A1
• Time in Patent Office: 957 Days
• Field of Search: None
• US Class Current: 718/101
• CPC Class Codes: G06F 2209/5013 Request control
• CPC Class Codes: G06F 9/466 Transaction processing
• CPC Class Codes: G06F 9/4843 by program, e.g. task dispatcher, supervisor, operating system

Abstract

In accordance with embodiments disclosed herein, there are provided mechanisms and methods for batch processing in an on-demand service environment. For example, in one embodiment, mechanisms include receiving a processing request for a multi-tenant database, in which the processing request specifies processing logic and a processing target group within the multi-tenant database. Such an embodiment further includes dividing or chunking the processing target group into a plurality of processing target sub-groups, queuing the processing request with a batch processing queue for the multi-tenant database among a plurality of previously queued processing requests, and releasing each of the plurality of processing target sub-groups for processing in the multi-tenant database via the processing logic at one or more times specified by the batch processing queue.

Claims

1. A method comprising:

receiving a processing request for a multi-tenant database, wherein the processing request specifies processing logic and a processing target group within the multi-tenant database;

dividing the processing target group into a plurality of processing target sub-groups;

queuing the processing request with a batch processing queue for the multi-tenant database among a plurality of previously queued processing requests;

releasing each of the plurality of processing target sub-groups for processing in the multi-tenant database via the processing logic at one or more times specified by the batch processing queue, the releasing being subject to one or more release limits enforced by a batch processing governor;

wherein violation of any of the one or more release limits causes the batch processing governor to throttle releases of batch processing jobs to the multi-tenant database; and

wherein the one or more release limits are selected from the group comprising;

(i) a maximum amount of memory permissibly allocated to each release, (ii) a maximum number of database statements permissibly executable with each release, (iii) a maximum number of permissible relational database row transactions for each release, (iv) a maximum number of permissible non-relational database object transactions for each release, (v) a maximum amount of database heap permissibly allocated or consumed by each release, (vi) a maximum total number of aggregate transactions by any one customer organization within a specified time period, (vii) a maximum total number of aggregate releases for batch processing jobs by any one customer organization within a specified time period, (viii) a maximum number of concurrent batch processing jobs releasable to the multi-tenant database, and (ix) a maximum current workload of the multi-tenant database.

2. The method of claim 1, wherein the processing request comprises a batch processing request having database transactions therein to be processed asynchronously in the multi-tenant database;

and wherein the method further comprises;

running each of the plurality of processing target sub-groups released for processing in the multi-tenant database concurrently with a plurality of other database transactions to be processed synchronously in the multi-tenant database.

3. The method of claim 1:

wherein the processing request specifies batch processing;

wherein the processing logic comprises customer logic to be applied to the processing target group in the multi-tenant database; and

wherein the customer logic is applied to each of the plurality of processing target sub-groups in the multi-tenant database when released by the batch processing queue.

4. The method of claim 1, wherein releasing each of the plurality of processing target sub-groups for processing at the one or more times specified by the batch processing queue comprises:

waiting until the queued processing request reaches a top position in the batch processing queue;

selecting an unprocessed one of the plurality of processing target sub-groups; and

releasing the selected unprocessed one of the plurality of processing target sub-groups to the multi-tenant database.

5. The method of claim 4, further comprising:

re-queuing the processing request with the batch processing queue when at least one of the plurality of processing target sub-groups for the processing request remain unprocessed; and

iteratively repeating the waiting, selecting, and releasing operations until all of the plurality of processing target sub-groups for the processing request are processed in the multi-tenant database.

6. The method of claim 1, wherein queuing the processing request with the batch processing queue comprises:

generating a reference message for the processing request in the batch processing queue, wherein the reference message specifies;

the one or more processing target sub-groups,

a processing state for each of the one or more processing target sub-groups,

a pointer to each of the one or more processing target sub-groups, and

a pointer to the processing logic to be applied to each of the one or more processing target sub-groups; and

wherein the batch processing queue manages a list of pending batch processing requests to be asynchronously processed against the multi-tenant database, each pending batch processing requests having one or more processing target sub-groups in an unprocessed state.

7. The method of claim 1:

wherein queuing the processing request with the batch processing queue comprises submitting a reference message for the processing request to the batch processing queue;

wherein releasing each of the plurality of processing target sub-groups for processing in the multi-tenant database comprises de-queuing the reference message from the batch processing queue; and

wherein the method further comprises submitting a new reference message for the processing request to the batch processing queue as part of a commit process for the respective processing target sub-group released for processing in the multi-tenant database, wherein the new reference message corresponds to a next unprocessed one of the plurality of processing target sub-groups for the processing request when at least one processing target sub-group for the processing request remains in an unprocessed state.

8. The method of claim 1:

wherein receiving the processing request comprises receiving the processing request at a host organization having the multi-tenant database operating therein; and

wherein the processing request is one of a plurality of requests received from a plurality of customer organizations, wherein each customer organization is an entity selected from the group consisting of;

a separate and distinct remote organization, an organizational group within the host organization, a business partner of the host organization, or a customer organization that subscribes to cloud computing services provided by the host organization.

9. The method of claim 1:

wherein the processing request is a batch processing request from a customer organization submitted to a host organization having the multi-tenant database operating therein;

wherein the size threshold is specified by the customer organization within the batch processing request; and

wherein dividing the processing target group into a plurality of processing target sub-groups comprises chunking the processing target group at the host organization based on the size threshold specified by the customer organization.

10. The method of claim 1, wherein the multi-tenant database comprises elements of hardware and software that are shared by a plurality of separate and distinct customer organizations, each of the separate and distinct customer organizations being remotely located from the host organization having the multi-tenant database operating therein.

11. The method of claim 1, wherein the processing target group within the multi-tenant database specified by the processing request comprises a plurality of rows within one or more tables of a relational data store in the multi-tenant database or a plurality of objects in a non-relational data store in the multi-tenant database, or both.

12. The method of claim 11:

wherein the relational data store comprises a relational database implementation selected from the group comprising;

an Oracle compatible database implementation, an IBM DB2 Enterprise Server compatible relational database implementation, a MySQL compatible relational database implementation, and a Microsoft SQL Server compatible relational database implementation; and

wherein the non-relational data store comprises a NoSQL non-relational database implementation selected from the group comprising a Vampire compatible non-relational database implementation, an Apache Cassandra compatible non-relational database implementation, a BigTable compatible non-relational database implementation, and an HBase compatible non-relational database implementation.

13. A method comprising:

receiving a processing request for a multi-tenant database, wherein the processing request specifies processing logic and a processing target group within the multi-tenant database;

dividing the processing target group into a plurality of processing target sub-groups;

queuing the processing request with a batch processing queue for the multi-tenant database among a plurality of previously queued processing requests;

releasing each of the plurality of processing target sub-groups for processing in the multi-tenant database via the processing logic at one or more times specified by the batch processing queue;

sending a completion notification to the originator of the processing request, wherein the completion notification comprises a completion report specifying;

(i) each of the plurality of processing target sub-groups released for processing in the multi-tenant database, and (ii) a completion status for each of the plurality of processing target sub-groups indicating a successful or unsuccessful completion for each processing target sub-group; and

wherein each completion status comprises a status selected from the group comprising;

(i) complete and committed without error, (ii) complete and committed after re-try due to excessive use of resources, (iii) complete and committed after re-try due to excessive multi-tenant database workload, and (iv) abort due to one or more errors for the respective processing target sub-group, wherein any database transaction error for the respective processing target sub-group prevents committing transaction associated with the respective processing target sub-group to the multi-tenant database.

14. A method comprising:

receiving a processing request for a multi-tenant database, wherein the processing request specifies processing logic and a processing target group within the multi-tenant database;

dividing the processing target group into a plurality of processing target sub-groups, wherein the dividing comprises (i) determining a total size of the processing target group, and (ii) dividing the processing target group into the one or more processing target sub-groups based on a size threshold, wherein the one or more processing target sub-groups do not overlap, and further wherein the one or more processing target sub-groups correspond in size to the determined total size;

queuing the processing request with a batch processing queue for the multi-tenant database among a plurality of previously queued processing requests;

releasing each of the plurality of processing target sub-groups for processing in the multi-tenant database via the processing logic at one or more times specified by the batch processing queue; and

wherein the total size of the processing target group corresponds to a criterion selected from the group comprising;

(i) a total number of records in the multi-tenant database to be processed in accordance with the processing request, (ii) a total number of rows in the multi-tenant database to be processed in accordance with the processing request, (iii) a total number of objects in the multi-tenant database to be processed in accordance with the processing request, and (iv) a total number of database statements to be transacted against the multi-tenant database in accordance with the processing request.

15. Non-transitory computer readable storage media having instructions stored thereon that, when executed by a system in a host organization having a processor and memory therein, the instructions cause the system to perform operations comprising:

receiving a processing request for a multi-tenant database, wherein the processing request specifies processing logic and a processing target group within the multi-tenant database;

dividing the processing target group into a plurality of processing target sub-groups;

queuing the processing request with a batch processing queue for the multi-tenant database among a plurality of previously queued processing requests;

releasing each of the plurality of processing target sub-groups for processing in the multi-tenant database via the processing logic at one or more times specified by the batch processing queue, the releasing being subject to one or more release limits enforced by a batch processing governor;

wherein violation of any of the one or more release limits causes the batch processing governor to throttle releases of batch processing jobs to the multi-tenant database; and

wherein the one or more release limits are selected from the group comprising;

(i) a maximum amount of memory permissibly allocated to each release, (ii) a maximum number of database statements permissibly executable with each release, (iii) a maximum number of permissible relational database row transactions for each release, (iv) a maximum number of permissible non-relational database object transactions for each release, (v) a maximum amount of database heap permissibly allocated or consumed by each release, (vi) a maximum total number of aggregate transactions by any one customer organization within a specified time period, (vii) a maximum total number of aggregate releases for batch processing jobs by any one customer organization within a specified time period, (viii) a maximum number of concurrent batch processing jobs releasable to the multi-tenant database, and (ix) a maximum current workload of the multi-tenant database.

16. The non-transitory computer readable storage medium of claim 15:

wherein the multi-tenant database comprises elements of hardware and software that are shared by a plurality of separate and distinct customer organizations, each of the separate and distinct customer organizations being remotely located from the host organization having the multi-tenant database operating therein; and

wherein the processing target group within the multi-tenant database specified by the processing request comprises a plurality of rows within one or more tables of a relational data store in the multi-tenant database or a plurality of objects in a non-relational data store in the multi-tenant database, or both.

17. A system comprising:

a processor and memory to execute instructions;

a request interface to receive a processing request for a multi-tenant database communicably interfaced with the system, wherein the processing request to specify processing logic and a processing target group within the multi-tenant database;

a request chunker to chunk the processing target group into a plurality of processing target sub-groups;

a batch processing queue to queue the processing request for release to the multi-tenant database; and

a release agent to release each of the plurality of processing target sub-groups for processing in the multi-tenant database via the processing logic;

a batch processing governor to enforce one or more release limits for releases by the release agent, wherein violation of any of the one or more release limits causes the batch processing governor to throttle releases of batch processing jobs to the multi-tenant database; and

wherein the one or more release limits are selected from the group comprising;

(i) a maximum amount of memory permissibly allocated to each release, (ii) a maximum number of database statements permissibly executable with each release, (iii) a maximum number of permissible relational database row transactions for each release, (iv) a maximum number of permissible non-relational database object transactions for each release, (v) a maximum amount of database heap permissibly allocated or consumed by each release, (vi) a maximum total number of aggregate transactions by any one customer organization within a specified time period, (vii) a maximum total number of aggregate releases for batch processing jobs by any one customer organization within a specified time period, (viii) a maximum number of concurrent batch processing jobs releasable to the multi-tenant database, and (ix) a maximum current workload of the multi-tenant database.