-
Notifications
You must be signed in to change notification settings - Fork 888
OPA is CPU inefficient #7438
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Yea, this has been an issue. Your global cache is a good solution. I spent a lot of time on OPA trying to reduce things, and have found about 1ms per request is what we could reasonably expect. Even if our policy was just a "return true". |
Gotcha. @kylecarbs is investigating a SQL-oriented approach since the needs of our system is pretty basic right now. We might as well merge #7439 as the work is done though. |
SQL RBAC solution? If so, that would make integrating into the data model easier for sure. |
@Emyrk we should sync on some ideas here. I'll ping you |
Looking at CPU usage on dogfood, OPA "topdown" takes 48% of our CPU. While agent metadata created ~5 rps per active workspace, and we should move reporting to a websocket, RBAC is consuming an excessive amount of CPU for the functionality we're getting.
profile.pb.gz
cc @Emyrk @coadler @kylecarbs
The text was updated successfully, but these errors were encountered: