yihua opened a new issue, #990:
URL: https://github.com/apache/hudi-rs/issues/990

   `Storage`'s read counters record bytes and fetch counts for base-file reads, 
but not how long requests take, how large they are, or how many are in flight 
at once. On object storage, a reader's throughput depends on whether it keeps 
requests in flight while it decodes, and today there is no way to see that from 
hudi-rs, nor to read the counters from outside the crate.
   
   Proposal: add per-request latency, size and concurrency (peak in flight, and 
time with at least one request outstanding) to the counters for every fetch on 
the read path, expose them per `FileGroupReader`, and report them from the 
file-group benchmark with one reader per task.
   


-- 
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]

Reply via email to