[
https://issues.apache.org/jira/browse/CASSANDRA-17243?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475344#comment-17475344
]
Andres de la Peña commented on CASSANDRA-17243:
-----------------------------------------------
{quote}Thinking more about it I am wondering whether it won't change a bit
behavior then and which branches to target for the fix.
{quote}
Good point. Changing the meaning of the streaming throughput properties will
mean that users don't updating their config will get a throughput ~4.6% lower
on streaming. I guess that might be a problem for users that have
experimentally fine-tuned those throughput values, although I'm not sure how
likely is that to happen.
Maybe we could apply the fix only to trunk, since that kind of performance
tuning should be repeated at least in new minors such as 4.1, given that new
releases can have performance differences.
> Fix BYTES_PER_MEGABIT in StreamManager
> --------------------------------------
>
> Key: CASSANDRA-17243
> URL: https://issues.apache.org/jira/browse/CASSANDRA-17243
> Project: Cassandra
> Issue Type: Bug
> Components: Consistency/Streaming
> Reporter: Ekaterina Dimitrova
> Assignee: Andres de la Peña
> Priority: Normal
> Fix For: 3.0.x, 3.11.x, 4.0.x, 4.x
>
>
> While working on CASSANDRA-15234 I noticed BYTES_PER_MEGABIT constant in the
> {code:java}
> StreamManager
> {code}
> class. It was introduced in CASSANDRA-16959.
> The current formula converts actually bytes to mebibits.
> The change needed for 3.0, 3.11 and 4.0(I am currently changing rate
> parameters to be in MiB/s for trunk as part of CASSANDRA-15234):
> {code:java}
> public static final double BYTES_PER_MEGABIT = (1000 * 1000) / 8; // from bits
> {code}
> CC [~adelapena]
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
