Flexible computing

Patent metadata

Field	Value
Publication number	US-12106149-B2
Application number	US-202318140086-A
Country	US
Kind code	B2
Filing date	Apr 27, 2023
Priority date	May 14, 2020
Publication date	Oct 1, 2024
Grant date	Oct 1, 2024

Abstract

Embodiments of the present disclosure may provide dynamic and fair assignment techniques for allocating resources on a demand basis. Assignment control may be separated into at least two components: a local component and a global component. Each component may have an active dialog with each other; the dialog may include two aspects: 1) a demand for computing resources, and 2) a total allowed number of computing resources. The global component may allocate resources from a pool of resources to different local components, and the local components in turn may assign their allocated resources to local competing requests. The allocation may also be throttled or limited at various levels.

First claim

What is claimed is:

1. A method comprising:
receiving query information from a first set of plurality of devices arranged in a first foreground cluster and a second set of plurality of devices arranged in a second foreground cluster associated with a data system, the query information including sources of queries requested by the first set of plurality of devices, the first foreground cluster being limited to handling only a first service type and the second foreground cluster being limited to handling only a second service type;
generating a first demand for computing resources for the data system based on the query information;
receiving a first distribution based on the first demand;
determining that a first source of queries has reached a cap limit of number of computing resources to be allocated for processing queries from the first source;
transmitting a first instruction to the first set of plurality of devices to release a first computing resource from the first distribution allocated to processing queries of the first service type from the first source based on determining that the first source has reached the cap limit;
transmitting a second instruction to the second set of plurality devices to release a second computing resource from the first distribution allocated to processing queries of the second service type from the first resource based on determining that the first source has reached the cap limit;
receiving a notification from the first of computing resources device regarding release of the first and second computing resources; and
allocating the released first and second computing resources to process queries from a different source.

2. The method of claim 1 , wherein the first demand for computing resources is based on a per source demand information.

3. The method of claim 1 , further comprising:
generating a second demand for computing resources based on historical data, wherein a second distribution of computing resources is based on the second demand for computing resources.

4. The method of claim 1 , wherein the first distribution of computing resources is based on a fair distribution algorithm.

5. The method of claim 4 , wherein the fair distribution algorithm includes a Max-Min fairness algorithm.

6. The method of claim 1 , further comprising:
transmitting the first demand for computing resources including writing a message in a metadata database that is accessible to a central server.

7. The method of claim 1 , further comprising:
transmitting the first demand for computing resources to a central server by via remote procedure calls.

8. The method of claim 1 , wherein the first demand for computing resources is based on consolidated query information received from the plurality of devices.

9. The method of claim 1 , wherein the computing resources comprises virtual computing resources.

10. A system comprising:
one or more processors of a machine; and
a memory storing instructions that, when executed by the one or more processors, cause the machine to perform operations comprising:
receiving query information from a first set of plurality of devices arranged in a first foreground cluster and a second set of plurality of devices arranged in a second foreground cluster associated with a data system, the query information including sources of queries requested by the first set of plurality of devices, the first foreground cluster being limited to handling only a first service type and the second foreground cluster being limited to handling only a second service type;
generating a first demand for computing resources for the data system based on the query information;
receiving a first distribution based on the first demand;
determining that a first source of queries has reached a cap limit of number of computing resources to be allocated for processing queries from the first source;
transmitting a first instruction to the first set of plurality of devices to release a first computing resource from the first distribution allocated to processing queries of the first service type from the first source based on determining that the first source has reached the cap limit;
transmitting a second instruction to the second set of plurality devices to release a second computing resource from the first distribution allocated to processing queries of the second service type from the first resource based on determining that the first source has reached the cap limit;
receiving a notification from the first of computing resources device regarding release of the first and second computing resources; and
allocating the released first and second computing resources to process queries from a different source.

11. The system of claim 10 , wherein the first demand for computing resources is based on a per source demand information.

12. The system of claim 10 , the operations further comprising:
generating a second demand for computing resources based on historical data, wherein a second distribution of computing resources is based on the second demand for computing resources.

13. The system of claim 10 , wherein the first distribution of computing resources is based on a fair distribution algorithm.

14. The system of claim 13 , wherein the fair distribution algorithm includes a Max-Min fairness algorithm.

15. The system of claim 10 , the operations further comprising:
transmitting the first demand for computing resources including writing a message in a metadata database that is accessible to a central server.

16. The system of claim 10 , the operations further comprising:
transmitting the first demand for computing resources to a central server by via remote procedure calls.

17. The system of claim 10 , wherein the first demand for computing resources is based on consolidated query information received from the plurality of devices.

18. The system of claim 10 , wherein the computing resources comprises virtual computing resources.

19. A non-transitory computer readable storage media storing instructions that, when executed by one or more processors, cause the one or more processors to:
receiving query information from a first set of plurality of devices arranged in a first foreground cluster and a second set of plurality of devices arranged in a second foreground cluster associated with a data system, the query information including sources of queries requested by the first set of plurality of devices, the first foreground cluster being limited to handling only a first service type and the second foreground cluster being limited to handling only a second service type;
generating a first demand for computing resources for the data system based on the query information;
receiving a first distribution based on the first demand;
determining that a first source of queries has reached a cap limit of number of computing resources to be allocated for processing queries from the first source;
transmitting a first instruction to the first set of plurality of devices to release a first computing resource from the first distribution allocated to processing queries of the first service type from the first source based on determining that the first source has reached the cap limit;
transmitting a second instruction to the second set of plurality devices to release a second computing resource from the first distribution allocated to processing queries of the second service type from the first resource based on determining that the first source has reached the cap limit;
receiving a notification from the fi

Assignees

• Snowflake Inc

Inventors

• Cruanes Thierry
• Demura Igor
• Ganesh Varun
• Rajaperumal Prasanna
• Wang Libo
• Yan Jiaqi

Classifications

• H04L67/1001

  for accessing one among a plurality of replicated servers · CPC title

• G06F9/5083

  Techniques for rebalancing the load in a distributed system · CPC title

• G06F9/542

  Event management; Broadcasting; Multicasting; Notifications · CPC title

• G06F2209/5011

  Pool · CPC title

• G06F9/5077

  Logical partitioning of resources; Management or configuration of virtualized resources (specific details on emulation or internal functioning of virtual machines G06F9/455) · CPC title