Starvationless kernel-aware distributed scheduling of software licenses

• US 9,792,419 B2
• Filed: 08/25/2014
• Issued: 10/17/2017
• Est. Priority Date: 08/11/2011
• Status: Expired due to Fees

• Current Assignee: Piece Future Pte., Ltd.
• Original Assignee: Alcatel-Lucent USA, Inc. (Nokia Corporation)
• Inventors: Carroll, Martin D., Katsak, William, Woithe, Hans C.
• Primary Examiner(s): TEETS, BRADLEY A
• Application Number:
  US14/467,488
• Publication Number:
  US 20140366162A1
• Time in Patent Office: 1,149 Days
• Field of Search: None
• CPC Class Codes: G06F 21/12 Protecting executable software
• CPC Class Codes: G06F 9/52 Program synchronisation; Mutual exclusion, e.g. by means of semaphores

Abstract

Methods, systems, and apparatuses for implementing shared-license software management are provided. Shared-license software management may be performed by generating a request for a license for running a process of a shared-license software application. The request is communicated to a license server, and the process is made available for scheduling an operation when a grant for the requested license is received from the license server, wherein the grant comprises a license quantum, which is an amount of central processing unit time for which the license has been granted to the process.

Claims

1. A system, comprising:

a memory; and

a processor in communication with the memory, the processor configured to;

execute a local scheduler as a hook into an existing kernel scheduler of a client device, wherein the local scheduler is configured to;

receive a set of license needs for running a process of a shared-license software application on a client device, the set of license needs comprising license needs served by different ones of a plurality of license servers;

partition the set of license needs into buckets, each bucket associated with one of the plurality of license servers, wherein each license need is partitioned into the bucket that is associated with the license server from which the license need is served;

determine a priority order for the buckets based on a priority of their associated license server;

generate requests for the license needs in each bucket according to the priority order such that a grant for a request for a license need in a particular bucket is received from its respective license server prior to generating a request for a license need in a next bucket in the priority order;

communicate each of the generated requests to its respective license server;

receive a respective grant for each of the generated requests from its respective license server, wherein the respective grant comprises a license quantum representing an amount of central processing unit (CPU) time which has been granted to the process; and

schedule the process of the shared-license software application for running on the client device when the respective grant for each of the generated requests is received and when the process of the shared-license software application is permitted to run based on semantics of the existing kernel-scheduler.

2. The system of claim 1, wherein the local scheduler is further configured to:

idle the process;

cancel at least one respective grant; and

add the process to a queue maintained for processes waiting for one or more respective grants.

3. The system of claim 1, wherein the local scheduler is further configured to decrement an amount of remaining license quantum for running the process by a quantum of time when the process has run for the quantum of time.

4. The system of claim 3, wherein the local scheduler is further configured to:

idle the process; and

add the process to a queue maintained for processes waiting for one or more respective grants if the amount of remaining license quantum is insufficient for running the process.

5. The system of claim 1, wherein the local scheduler is further configured to generate the requests based on the set of license needs.

6. The system of claim 1, wherein the local scheduler is further configured to determine whether the process is one of expired, killed, idle or nonidle.

7. The system of claim 1, wherein the local scheduler is further configured to determine whether the process is idle for longer than a predetermined period of time.

8. The system of claim 1, wherein the local scheduler is further configured to cancel at least one respective grant.

9. The system of claim 1, wherein the local scheduler is further configured to perform one or more operations via a kernel scheduler.

10. The system of claim 1, wherein the local scheduler is further configured to estimate a license quantum consumption rate for the process.

11. The system of claim 1, wherein the requests for license needs are generated such that requests for license needs of a bucket having a higher priority in the priority order are generated prior to requests for license needs of a bucket having a lower priority in the priority order.

12. A method comprising:

executing a local scheduler as a hook into an existing kernel scheduler of a client device;

receiving, by the local scheduler, a set of license needs for running a process of a shared-license software application on a client device, the set of license needs comprising license needs served by different ones of a plurality of license servers;

partitioning, by the local scheduler, the set of license needs into buckets, each bucket associated with one of the plurality of license servers, wherein each license need is partitioned into the bucket that is associated with the license server from which the license need is served;

determining, by the local scheduler, a priority order for the buckets based on a priority of their associated license server;

generating, by the local scheduler, requests for the license needs in each bucket according to the priority order such that a grant for a request for a license need in a particular bucket is received from its respective license server prior to generating a request for a license need in a next bucket in the priority order;

communicating, by the local scheduler, each of the generated requests to its respective license server;

receiving, by the local scheduler, a respective grant for each of the generated requests from its respective license server, wherein the respective grant comprises a license quantum representing an amount of central processing unit (CPU) time which has been granted to the process; and

scheduling, by the local scheduler, the process of the shared-license software application for running on the client device when the respective grant for each of the generated requests is received and when the process of the shared-license software application is permitted to run based on semantics of the existing kernel-scheduler.

13. The method of claim 12, further comprising:

idling the process;

canceling at least one respective grant; and

adding the process to a queue maintained for processes waiting for one or more respective grants.

14. The method of claim 12, further comprising decrementing an amount of remaining license quantum for running the process by a quantum of time when the process has run for the quantum of time.

15. The method of claim 14, further comprising:

idling the process; and

adding the process to a queue maintained for processes waiting for one or more respective grants if the amount of remaining license quantum is insufficient for running the process.

16. The method of claim 12, wherein generating the requests comprises generating the requests based on the set of license needs.

17. The method of claim 12, further comprising determining whether the process is one of expired, killed, idle or nonidle.

18. The method of claim 12, further comprising determining whether the process is idle for longer than a predetermined period of time.

19. The method of claim 12, further comprising canceling at least one respective grant.

20. The method of claim 12, further comprising performing one or more operations via a kernel scheduler.

21. The method of claim 12, further comprising estimating a license quantum consumption rate for the process.

22. The method of claim 12, wherein the requests for license needs are generated such that requests for license needs of a bucket having a higher priority in the priority order are generated prior to requests for license needs of a bucket having a lower priority in the priority order.

23. An article of manufacture including a non-transitory computer-readable medium having instructions stored thereon, that in response to execution by a computing device causes the computing device to perform operations comprising:

executing a local scheduler as a hook into an existing kernel scheduler of a client device;

receiving, by the local scheduler, a set of license needs for running a process of a shared-license software application on a client device, the set of license needs comprising license needs served by different ones of a plurality of license servers;

partitioning, by the local scheduler, the set of license needs into buckets, each bucket associated with one of the plurality of license servers, wherein each license need is partitioned into the bucket that is associated with the license server from which the license need is served;

determining, by the local scheduler, a priority order for the buckets based on a priority of their associated license server;

generating, by the local scheduler, requests for the license needs in each bucket according to the priority order such that a grant for a request for a license need in a particular bucket is received from its respective license server prior to generating a request for a license need in a next bucket in the priority order;

communicating, by the local scheduler, each of the generated requests to its respective license server;

receiving, by the local scheduler, a respective grant for each of the generated requests from its respective license server, wherein the respective grant comprises a license quantum representing an amount of central processing unit (CPU) time which has been granted to the process; and

scheduling, by the local scheduler, the process of the shared-license software application for running on the client device when the respective grant for each of the generated requests is received and when the process of the shared-license software application is permitted to run based on semantics of the existing kernel-scheduler.