Query Offloading Runtime for Database Compute Spikes

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing database systems face challenges in efficiently handling compute-intensive workloads due to unpredictable query spikes, resource wastage, and inefficient offloading of subqueries, leading to suboptimal performance and cost inefficiencies.

Innovation Solution

A compute-offloading framework that dynamically provisions and offloads compute-intensive workloads to a heterogeneous set of machines, utilizing pipeline execution, semi-stateful execution, and fine-grained data distribution to optimize resource usage and performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the database cluster is statically provisioned to handle query spikes, then the system can handle peak workloads, but resources are wasted during low-demand periods

Engineering Contradiction:

Improveability to handle query spikesVSAvoidresource wastage

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The patent implements dynamic cluster provisioning where the number of database nodes is adjusted based on actual workload demands. The system automatically adds nodes during query spikes and removes nodes during low-demand periods, transitioning from static to dynamic cluster sizing. This resolves the contradiction by enabling the system to handle peak workloads when needed while eliminating resource wastage during off-peak times.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes the parameter of cluster size dynamically based on workload patterns. By monitoring query volumes and adjusting the number of active nodes accordingly, the system optimizes resource utilization. This parameter change approach allows the cluster to scale up during spikes and scale down during quiet periods, balancing reliability and resource efficiency.

Inventive Principle:

Principle #35Parameter changes

2Reliability

If the database cluster size is increased to handle spikes, then query execution can maintain performance during peaks, but query execution becomes unnecessarily slow during normal operations due to resource over-provisioning

Engineering Contradiction:

Improvequery performance during spikesVSAvoidquery execution speed during normal operations

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system dynamically adjusts cluster size based on actual query workload. During normal operations with low demand, the cluster shrinks to a smaller size, enabling faster query execution by reducing resource contention and overhead. During spikes, the cluster expands to maintain performance. This dynamic adjustment resolves the contradiction between maintaining peak performance and optimizing normal operation speed.

Inventive Principle:

Principle #15Dynamics

3Adaptability or versatility

If new nodes are added to the cluster during runtime, then the cluster can handle increased workload, but the new nodes are only picked up by future queries and not currently executing queries

Engineering Contradiction:

Improvedynamic cluster scalingVSAvoidquery execution time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-warming new nodes and pre-positioning them in the cluster before they are needed. When a node is added, it is prepared in advance and registered with the query executor, so that when workload increases, the new node can immediately participate in currently executing queries without delay. This preliminary preparation enables both dynamic scaling and immediate utilization of new nodes.

Inventive Principle:

Principle #10Preliminary action

4Ease of manufacture

If the database cluster uses homogenous machine types, then hardware provisioning is simplified, but the system cannot optimize for different compute patterns and workload types

Engineering Contradiction:

Improvehardware provisioning simplicityVSAvoidcompute optimization

Core Design Contradiction:

Ease of manufactureVSProductivity

Solution Approach 1:

The patent introduces heterogeneity at the node level, allowing different machine types to be deployed within the cluster based on specific compute requirements. Instead of uniform homogenous nodes, the system can allocate different hardware configurations to different query types or workload patterns. This local quality differentiation enables optimized compute resource allocation while maintaining overall cluster manageability.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20260072760A1System And Method To Scale Out For Compute-Intensive Workloads

Publication Date: 2026.03.12 ORACLE INT CORP

AI summary

A method and apparatus for offloading compute-intensive workloads is provided. A database system compiles an execution plan to generate an offload-enabled plan by identifying a candidate offloading region in the execution plan, generating and adding an offloading branch in the offload-enabled plan, corresponding to the candidate offloading region, for execution by a compute offload runtime, wherein the compute offload runtime comprises a compute offload runtime library executing on the database system and on each node of a compute offload server, and adding the candidate offloading region as a fallback branch in the offload-enabled plan. The database system executes the offload-enabled plan by executing the offloading branch using one or more compute nodes in the database server or the compute offload server using the offload runtime or by executing the fallback branch using one or more compute nodes in the database server.