[ 
https://issues.apache.org/jira/browse/HADOOP-12409?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14745056#comment-14745056
 ] 

Steve Loughran commented on HADOOP-12409:
-----------------------------------------

having separate MR/YARN JIRAs will result them all being tagged as fixed —what 
it does do is (a) let people on the mapreduce-dev and yarn-dev lists see that 
you are proposing changes to their code and (b) have jenkins test those bits of 
the code with the patch. Changing those bits of the code without the warnings 
or the tests will only upset the teams and lead to patches being reverted

Clock-wise, how about adding a new method, `monotonicTimeMillis()`, which is 
just nanoTime/1e6; easy to switch from one to the other.

In a followup JIRA we could look at switching some of those loops off their own 
(repeated) delay logic and move to some re-usable duration measuring code. I 
have some nearby 
[https://github.com/apache/incubator-slider/blob/develop/slider-core/src/main/java/org/apache/slider/common/tools/Duration.java],
 which could act as a starting point

> Move org.apache.hadoop.yarn.util.Clock and relative implementations to hadoop 
> Common
> ------------------------------------------------------------------------------------
>
>                 Key: HADOOP-12409
>                 URL: https://issues.apache.org/jira/browse/HADOOP-12409
>             Project: Hadoop Common
>          Issue Type: Improvement
>          Components: util
>    Affects Versions: 3.0.0
>            Reporter: Xianyin Xin
>            Assignee: Xianyin Xin
>            Priority: Minor
>         Attachments: Hadoop-12409.001.patch, Hadoop-12409.002.patch
>
>
> It is widely used by MR and YARN.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to