-
Notifications
You must be signed in to change notification settings - Fork 2.1k
Adds metric tags to storage instrumentation #9595
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
Conversation
765ed7d to
019398a
Compare
|
@ReubenBond, could I trouble you to have a look at this PR? |
|
Thanks for the ping, @cdemi - I left a comment. I believe we should address that before merge |
2587cb3 to
6d8bd0a
Compare
|
Thanks for the feedback, @ReubenBond! I have made changes according to your feedback. I can't pass |
|
Thanks, @cdemi! I added one last comment, otherwise LGTM |
578736c to
198f1b5
Compare
|
You're absolutely right. Should be OK now |
198f1b5 to
8a11fb7
Compare
It's clearer that we are referring to the type name of the provider and not the user-defined name given to the provider during configuration
8a11fb7 to
36dd782
Compare
This PR aims to add support for tagging telemetry with contextual metadata (Storage Provider Name, State Name and State Type).
The current telemetry for storage operations (read/write/delete/errors) provides aggregate insights but lacks the granularity needed to troubleshoot issues or analyze behavior per provider or state.
This will enable, for example, Prometheus queries like:
I think it will be very useful for identifying noisy or error-prone storage providers, differentiating latency by grain state type or maybe alerting on specific grain state issues.
The
StateStorageBridge<TState>constructs the tag array once per instance and reuses it for all metric operations. Since I made changes here I also changed the storage provider type name.GetType().Nameto be computed once during construction instead of on every operation, reducing redundant string allocationsMicrosoft Reviewers: Open in CodeFlow