Xuanwo opened a new issue, #5878: URL: https://github.com/apache/opendal/issues/5878
It's a long-term issue that the metrics generated by OpenDAL lack guidance. That said, OpenDAL generated many metrics, but users still don't know how to use them or what they mean. For examples: - Users don't understand the differences between `reader::start` and `reader::read` - Users often ask how to calculate time to first byte. (it's `reader::start`, but no body knows) I have tested our newly [refactored metrics system](https://github.com/apache/opendal/issues/5788) in applications and spoken with some project users, and they still have the same feedback. This confirms that we are heading in the wrong direction—users want clearer, more direct metrics that can answer their questions explicitly. The newly added https://github.com/apache/opendal/issues/5794 is a good start but not enough. So I want to add the following changes to our metrics, especially for read and write: - Change `operation_duration_seconds { operation: read }` to entire read operation time. - Add `operation_executing` to reflect the operations that is executing. - Change existing `operation_bytes { operation: read }` to entire read operation bytes. - Add `operation_bytes_total` to reflects the total bytes - Add `operation_count_total` to reflects operations times, users don't need to calculate them by `operation_duration_seconds` anymore. - Add `operation_bytes_rate` to reflects the bytes rate distribution of operations. - Add `operation_ttfb_seconds` to reflects the time to first byte of operations. - Provide a better default value for our histogram buckets. Welcome to leave your feedback over this change! -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
