Amazon Web Services, part of Amazon.com Inc. (NASDAQ: AMZN), has updated the storage layer behind Amazon Aurora DSQL, which AWS said can raise write throughput by up to 50% for many transactional workloads. The change is available to new and existing clusters in every AWS Region where Aurora DSQL operates, giving database teams a capacity improvement without a stated migration or added charge.
The update targets workloads that write frequently while preserving an ordered primary key, including audit logs, ledgers and event feeds. AWS said tables keyed by timestamps, identity columns, ULIDs or UUIDv7 identifiers can see larger gains. Those patterns let applications retain time-based ordering for reads while accommodating a heavier stream of inserts.
For IT operations teams, audit and event data can move quickly when an application, service or security control records activity at high volume. AWS said an audit-log table keyed by event time processed up to 200% more inserts per second in its tests. The company did not publish the hardware configuration, dataset size, baseline throughput or test duration, so that result should be read as an AWS test result rather than a general production benchmark.
The central change is in the service’s storage layer rather than a new application interface or database feature. That matters because the company says existing clusters receive the improvement alongside new ones. Organizations running write-heavy DSQL workloads may be able to evaluate the new behavior with their current schemas before deciding whether to alter application design or capacity plans.
AWS specifically highlighted increasing-key tables. Timestamp- and sequence-based keys are common where records must be retrieved in chronological order, such as operational events, financial entries and compliance trails. The company said the update allows customers to keep that time-ordered design for write-heavy tables while completing data loads sooner. It did not specify the workload mix required to reach the stated 50% increase.
Availability across Aurora DSQL regions also gives distributed IT teams a consistent rollout boundary: the change applies where the service is already available, not to an announced future region set. AWS said it is included at no additional cost. The company directed customers to the Aurora DSQL User Guide for service details.
The improvement is a performance update, not a claim that all transactional databases will produce the same result. Teams will still need to measure insert rates, key patterns and latency against their own applications. But for organizations using Aurora DSQL for ordered, write-intensive operational records, the storage-layer change could reduce a constraint that otherwise forces a trade-off between chronological access and ingestion speed.
Legacy database performance projects often require a schema redesign, a service migration or additional infrastructure to increase write capacity. AWS is positioning this change as an in-place service improvement for existing clusters. Whether the gain changes a particular workload will depend on the table design and traffic pattern, but the stated support for time-ordered keys makes the update most relevant to operations teams managing large event and audit data flows.
It also gives database teams a defined service change to include in routine testing, capacity reviews and service-level monitoring, rather than treating the published figure as a substitute for workload-specific validation.
