-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45449/#review126148
-----------------------------------------------------------




geode-core/src/test/java/com/gemstone/gemfire/cache/client/internal/pooling/ConnectionManagerJUnitTest.java
 (line 300)
<https://reviews.apache.org/r/45449/#comment189049>

    Would it be better for start to be called right before the first 
"returnConnection"?
    If you look at line 272 you will see that someone recently moved start to 
be set right before the passivate. Passivate ends up calling System.nanoTime 
and setting that as the last accessed time. But it calls it before doing a sync 
and it may lose some time in the sync.
    
    In this call start is being set AFTER 4 returnConnection calls (which all 
call passivate) so the first two returned can expire earlier than the last two.
    
    By having an earlier "start" it makes it more likely that your assertion 
will pass since it wants elapsed to be >= the configured idle timeout.


- Darrel Schneider


On March 29, 2016, 3:08 p.m., Hitesh Khamesra wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/45449/
> -----------------------------------------------------------
> 
> (Updated March 29, 2016, 3:08 p.m.)
> 
> 
> Review request for geode, Bruce Schuchardt, Jianxia Chen, and Udo Kohlmeyer.
> 
> 
> Repository: geode
> 
> 
> Description
> -------
> 
> Aligned test code with ConnectionManager..
> 
> 
> Diffs
> -----
> 
>   
> geode-core/src/test/java/com/gemstone/gemfire/cache/client/internal/pooling/ConnectionManagerJUnitTest.java
>  bb407a4 
> 
> Diff: https://reviews.apache.org/r/45449/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Hitesh Khamesra
> 
>

Reply via email to