Predicate evaluation based on CPU utilization of shared storage

Patent metadata

Field	Value
Publication number	US-12373320-B1
Application number	US-202418769172-A
Country	US
Kind code	B1
Filing date	Jul 10, 2024
Priority date	Jul 10, 2024
Publication date	Jul 29, 2025
Grant date	Jul 29, 2025

Abstract

A hybrid transactional/analytical processing (HTAP) database includes an online transaction processing (OLTP) engine, an online analytical processing (OLAP) engine provided separately from the OLTP engine, and a shared storage holding data of multiple tenants. The shared storage is configured to receive a request from the OLAP engine for evaluation of a predicate against the data stored on the shared storage. The shared storage is further configured to determine that a CPU utilization of the shared storage exceeds a predicate refusal threshold. Upon determining that the CPU utilization of the shared storage exceeds the predicate refusal threshold, the shared storage is further configured to refrain from evaluating at least a portion of the predicate.

First claim

The invention claimed is:

1. A hybrid transactional/analytical processing (HTAP) database, comprising:
an online transaction processing (OLTP) engine;
an online analytical processing (OLAP) engine provided separately from the OLTP engine; and
a shared storage holding data of multiple tenants, wherein the shared storage is configured to:
receive a request from the OLAP engine for evaluation of a predicate against the data stored on the shared storage;
determine that a CPU utilization of the shared storage exceeds a predicate refusal threshold; and
upon determining that the CPU utilization of the shared storage exceeds the predicate refusal threshold, refrain from evaluating at least a portion of the predicate.

2. The HTAP database of claim 1 , wherein the shared storage is further configured to:
return a response to the request including partial or full unprocessed results from the data for predicate evaluation; and
attach partial or full predicate refusal metadata to the response indicating the partial or full unprocessed results have not been evaluated.

3. The HTAP database of claim 2 , wherein the OLAP engine is configured to:
receive the response including the partial or full unprocessed results from the data for predicate evaluation and the partial or full predicate refusal metadata;
determine that at least the portion of the predicate has not been evaluated based on the partial or full predicate refusal metadata; and
evaluate the predicate using the partial or full unprocessed results received in the response.

4. The HTAP database of claim 1 , wherein
the predicate refusal threshold is a partial predicate refusal threshold,
the share storage further determines a full predicate refusal threshold that is higher than the partial predicate refusal threshold, and
the shared storage is further configured to:
determine that the CPU utilization of the shared storage exceeds the partial predicate refusal threshold but does not reach the full predicate refusal threshold;
upon determining that the CPU utilization of the shared storage exceeds the partial predicate refusal threshold but does not reach the full predicate refusal threshold, refrain from evaluating a portion of the predicate;
return a response to the request including partial unprocessed results from the data for predicate evaluation; and
attach partial predicate refusal metadata to the response indicating the portion of the predicate has not been evaluated.

5. The HTAP database of claim 4 , wherein the OLAP engine is configured to:
receive the response including the partial unprocessed results from the data for predicate evaluation and the partial predicate refusal metadata;
determine that the portion of the predicate has not been evaluated based on the partial predicate refusal metadata; and
evaluate the predicate using the partial unprocessed results received in the response.

6. The HTAP database of claim 1 , wherein
the predicate refusal threshold is a partial predicate refusal threshold,
the share storage further determines a full predicate refusal threshold that is higher than the partial predicate refusal threshold, and
the shared storage is further configured to:
determine that the CPU utilization of the shared storage reaches the full predicate refusal threshold;
upon determining that the CPU utilization of the shared storage reaches the full predicate refusal threshold, refrain from evaluating the predicate;
return a response to the request including full unprocessed results from the data for predicate evaluation; and
attach full predicate refusal metadata to the response indicating the predicate has not been evaluated.

7. The HTAP database of claim 6 , wherein the OLAP engine is configured to:
receive the response including the full unprocessed results from the data for predicate evaluation and the full predicate refusal metadata;
determine that the predicate has not been evaluated based on the full predicate refusal metadata; and
evaluate the predicate using the full unprocessed results received in the response.

8. The HTAP database of claim 1 , wherein the shared storage is further configured to:
determine that the CPU utilization of the shared storage does not exceed the predicate refusal threshold;
upon determining that the CPU utilization of the shared storage does not exceed the predicate refusal threshold, evaluate the predicate at the shared storage; and
return a response to the request including full processed results from the data for predicate evaluation.

9. A data processing method for a hybrid transactional/analytical processing (HTAP) database including an online transaction processing (OLTP) engine, an online analytical processing (OLAP) engine provided separately from the OLTP engine, and a shared storage holding data of multiple tenants, the method comprising, at the shared storage:
receiving a request from the OLAP engine for evaluation of a predicate against the data stored on the shared storage;
determining that a CPU utilization of the shared storage exceeds a predicate refusal threshold; and
upon determining that the CPU utilization of the shared storage exceeds the predicate refusal threshold, refraining from evaluating at least a portion of the predicate.

10. The data processing method of claim 9 , further comprising, at the shared storage:
returning a response to the request including partial or full unprocessed results from the data for predicate evaluation; and
attaching partial or full predicate refusal metadata to the response indicating the partial or full unprocessed results have not been evaluated.

11. The data processing method of claim 10 , further comprising, at the OLAP engine:
receiving the response including the partial or full unprocessed results from the data for predicate evaluation and the partial or full predicate refusal metadata;
determining that at least the portion of the predicate has not been evaluated based on the partial or full predicate refusal metadata; and
evaluating the predicate using the partial or full unprocessed results received in the response.

12. The data processing method of claim 9 , wherein
the predicate refusal threshold is a partial predicate refusal threshold,
the share storage further determines a full predicate refusal threshold that is higher than the partial predicate refusal threshold, and
the method further comprises, at the shared storage:
determining that the CPU utilization of the shared storage exceeds the partial predicate refusal threshold but does not reach the full predicate refusal threshold;
upon determining that the CPU utilization of the shared storage exceeds the partial predicate refusal threshold but does not reach the full predicate refusal threshold, refraining from evaluating a portion of the predicate;
returning a response to the request including partial unprocessed results from the data for predicate evaluation; and
attaching partial predicate refusal metadata to the response indicating the portion of the predicate has not been evaluated.

13. The data processing method of claim 12 , further comprising, at the OLAP engine:
receiving the response including the partial unprocessed results from the data for predicate evaluation and the partial predicate refusal metadata;
determining that the portion of the predicate has not been evaluated based on the partial predicate refusal metadata; and
evaluating the predicate using the partial unprocessed results received in the response.

14. The data processing method of claim 9 , wherein
the predicate refusal threshold is a partial predicate refusa

Assignees

• Lemon Inc
• Douyin Vision Co Ltd

Inventors

• Ding Wei
• Qian Jianfeng
• Zeng Shicai
• Zhang Li
• Chen Jianjun

Classifications

• G06F2201/81

  Threshold · CPC title

• G06F2201/80

  Database-specific techniques · CPC title

• G06F16/283 Primary

  Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP · CPC title

• G06F11/3024 Primary

  where the computing system component is a central processing unit [CPU] · CPC title