[ 
https://issues.apache.org/jira/browse/CASSANDRA-12661?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15966159#comment-15966159
 ] 

Edward Capriolo commented on CASSANDRA-12661:
---------------------------------------------

I pushed some of the changes requested here:

https://github.com/apache/cassandra/compare/trunk...edwardcapriolo:CASSANDRA-12661-2?expand=1

I do not quite understand the defect causing the test failures. Side effect or 
stale file?

{quote}
due to the checks in the implementation, the order of calls to 
setGcLogThresholdInMs + setGcWarnThresholdInMs depend on the current values, 
which should not be the case. this could be changed by having just one method 
that changes both values.
{quote}

That surely reduces the bounds check logic. I would rather understand first why 
I can not reproduce the test failure. If the approach of setting individually  
is fundamentally flawed I suggest rolling this back.

> Make gc_log and gc_warn settable at runtime
> -------------------------------------------
>
>                 Key: CASSANDRA-12661
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-12661
>             Project: Cassandra
>          Issue Type: New Feature
>            Reporter: Edward Capriolo
>            Assignee: Edward Capriolo
>            Priority: Minor
>         Attachments: Screen Shot 2017-04-12 at 11.38.06 AM.png
>
>
> Changes:
> * Move gc_log_threshold_in_ms and gc_warn_threshold_in_ms close together in 
> the config
> * rename variables to match properties
> * add unit tests to ensure hybration
> * add unit tests to ensure variables are set propertly
> * minor perf (do not consturct string from buffer f not logging)



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Reply via email to