Skip to content
This repository has been archived by the owner on Feb 26, 2021. It is now read-only.

Sockets benchmarks #30

Open
divega opened this issue Feb 14, 2018 · 0 comments
Open

Sockets benchmarks #30

divega opened this issue Feb 14, 2018 · 0 comments
Assignees

Comments

@divega
Copy link

divega commented Feb 14, 2018

@roji is now writing some benchmarks that don't actually use a database driver as a means to rule out low-level bottlenecks as the root cause for the differences we are seeing with .NET vs. JDBC, Linux vs. Windows and sync vs. async in Npgsql results. We wonder if other teams have done this already, but it seems to have enough value to just do it again. If we are able to confirm the presence of low-level bottlenecks we can pass this test to someone else to do the profiling and analysis.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants