[
https://issues.apache.org/jira/browse/YARN-4712?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15175980#comment-15175980
]
Junping Du commented on YARN-4712:
----------------------------------
Back on this patch, it seems more things need to be fixed for UNAVAILABLE case,
like code blow:
{code}
// Multiply by 1000 to avoid losing data when converting to int
int milliVcoresUsed = (int) (cpuUsageTotalCoresPercentage * 1000
* maxVCoresAllottedForContainers /nodeCpuPercentageForYARN);
{code}
It sounds weird if cpuUsageTotalCoresPercentage is -1 in UNAVAILABLE case.
In addition, code below sounds not right:
{code}
+ cpuMetric.addValue(currentTimeMillis,
+ (long) Math.round(cpuUsageTotalCoresPercentage));
{code}
it make cpu metric to be either 0 or 1 which is not expected here?
> CPU Usage Metric is not captured properly in YARN-2928
> ------------------------------------------------------
>
> Key: YARN-4712
> URL: https://issues.apache.org/jira/browse/YARN-4712
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: timelineserver
> Reporter: Naganarasimha G R
> Assignee: Naganarasimha G R
> Labels: yarn-2928-1st-milestone
> Attachments: YARN-4712-YARN-2928.v1.001.patch,
> YARN-4712-YARN-2928.v1.002.patch
>
>
> There are 2 issues with CPU usage collection
> * I was able to observe that that many times CPU usage got from
> {{pTree.getCpuUsagePercent()}} is
> ResourceCalculatorProcessTree.UNAVAILABLE(i.e. -1) but ContainersMonitor do
> the calculation i.e. {{cpuUsageTotalCoresPercentage = cpuUsagePercentPerCore
> /resourceCalculatorPlugin.getNumProcessors()}} because of which UNAVAILABLE
> check in {{NMTimelinePublisher.reportContainerResourceUsage}} is not
> encountered. so proper checks needs to be handled
> * {{EntityColumnPrefix.METRIC}} uses always LongConverter but
> ContainerMonitor is publishing decimal values for the CPU usage.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
