Open WenyXu opened 3 months ago
Hi @WenyXu , I would like to work on this. Can you provide more insight into the problem?
Hi @WenyXu , I would like to work on this. Can you provide more insight into the problem?
Hey, I suggest starting with our tracing tutorials. Run a query and check out the traces—you'll probably notice they’re a bit too coarse. For example, we don't really see how much time is spent on things like downloading files or building readers. This issue is about improving that by adding more detailed, debug-level traces in the Mito engine, especially for queries.
What type of enhancement is this?
Other
What does the enhancement do?
GreptimeDB exports all collected spans using the gRPC-based OTLP protocol (see also).
However, current traces are still not good enough for performance debugging, consider adding more detailed spans in the mito engine. e.g., The time reading bytes from S3, The time building reader, etc.
Implementation challenges
No response