[ 
https://issues.apache.org/jira/browse/CASSANDRA-16325?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17656990#comment-17656990
 ] 

Isaac Reath commented on CASSANDRA-16325:
-----------------------------------------

Yes [~dcapwell]  this seems to be solving something very similar to what's 
happening in StreamingState::streamProgress in your patch. Looking at these 
lines in your patch: 
[https://github.com/apache/cassandra/pull/2066/files#diff-2c11c59ea199afd7097d89eece180d6ce6f9ee7568ffb141994f5dff9d9c2d27R267-R269]
 it looks very similar to the approach I initially took in this patch (see: 
[https://github.com/apache/cassandra/pull/2058#discussion_r1061577063).] We 
changed our approach to use this different one to avoid needing to keep track 
of the incoming / outgoing ProgressInfo objects (per [~smiklosovic]'s 
concerns). I think if we updated the ProgressInfo object to store the delta 
from the last time progress was called, then we could avoid having to track 
ProgressInfo objects and still leverage the ProgressEvent. From what I can 
tell, we have that information most of the time we call StreamSession.progress.

The only thought I have about moving the counters into the StreamingSession 
class is if there is a feature flag around per-file progress. As a user, I 
would still want metrics to report if that flag is enabled. 

 

> Update streaming metrics incrementally
> --------------------------------------
>
>                 Key: CASSANDRA-16325
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-16325
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Observability/Metrics
>            Reporter: Paulo Motta
>            Assignee: Isaac Reath
>            Priority: Normal
>              Labels: lhf
>             Fix For: 4.2
>
>          Time Spent: 10h 10m
>  Remaining Estimate: 0h
>
> Currently the inbound and outbound streamed bytes metrics are incremented 
> after each file is streamed, what doesn't represent the current number of 
> bytes streamed since it can take a long time for a large file to be streamed. 
> We should update the metric incrementally as data is streamed.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org

Reply via email to