Skip to content

Conversation

@Dav1dde
Copy link
Member

@Dav1dde Dav1dde commented Nov 28, 2025

Separate PR to have separate revisions for additional performance tests. ahash should just be an upgrade, for blocking our actual workload seems to favor the blocking mode over the default incremental.

Blocking resizes tend to be better in terms of throughput, especially in setups with multiple writers that can perform the resize in parallel. However, they can lead to latency spikes for insert operations that have to resize large tables.

Latency spikes on inserts are acceptable and we do have a lot of writers.

@Dav1dde Dav1dde self-assigned this Nov 28, 2025
@Dav1dde Dav1dde requested a review from a team as a code owner November 28, 2025 11:25
Base automatically changed from dav1d/opportunistic-quotas to master December 1, 2025 11:10
@Dav1dde Dav1dde force-pushed the dav1d/tune-quota-cache-hashmap branch from 5fd1133 to 5697730 Compare December 5, 2025 09:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants