hadoop-yarn-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sihai Ke (JIRA)" <j...@apache.org>
Subject [jira] [Created] (YARN-9550) Suspect wrong way to calculater container utilized vcore.
Date Tue, 14 May 2019 04:10:00 GMT
Sihai Ke created YARN-9550:
------------------------------

             Summary: Suspect wrong way to calculater container utilized vcore.
                 Key: YARN-9550
                 URL: https://issues.apache.org/jira/browse/YARN-9550
             Project: Hadoop YARN
          Issue Type: Bug
          Components: nodemanager
    Affects Versions: 2.9.1
            Reporter: Sihai Ke


In hadoop 2.9.1 class *ContainersMonitorImpl* line 664, I suspect it use the wrong way to
calculate the milliVcoresUsed, below is the code.
{code:java}
ResourceCalculatorProcessTree pTree = ptInfo.getProcessTree();
pTree.updateProcessTree();    // update process-tree
if (!pTree.isValidData()) {
  // If we cannot get the data for one container, we ignore it all
  LOG.error("Cannot get the data for " + pId);
  trackedContainersUtilization = null;
  continue;
}
long currentVmemUsage = pTree.getVirtualMemorySize();
long currentPmemUsage = pTree.getRssMemorySize();

// if machine has 6 cores and 3 are used,
// cpuUsagePercentPerCore should be 300% and
// cpuUsageTotalCoresPercentage should be 50%
float cpuUsagePercentPerCore = pTree.getCpuUsagePercent();
if (cpuUsagePercentPerCore < 0) {
  // CPU usage is not available likely because the container just
  // started. Let us skip this turn and consider this container
  // in the next iteration.
  LOG.info("Skipping monitoring container " + containerId
      + " since CPU usage is not yet available.");
  continue;
}

float cpuUsageTotalCoresPercentage = cpuUsagePercentPerCore /
    resourceCalculatorPlugin.getNumProcessors();

// Multiply by 1000 to avoid losing data when converting to int
int milliVcoresUsed = (int) (cpuUsageTotalCoresPercentage * 1000
    * maxVCoresAllottedForContainers /nodeCpuPercentageForYARN);

// milliPcoresUsed = (int) (cpuUsagePercentPerCore * 1000 / 100;
// As cpuUsagePercentagePerCore use 100 to represent 1 single core.
int milliPcoresUsed = (int) (cpuUsagePercentPerCore * 10);
// as processes begin with an age 1, we want to see if there
// are processes more than 1 iteration old.
vcoresUsageByAllContainers += milliVcoresUsed;
pcoresByAllContainers += milliPcoresUsed;
{code}
 

I think

 
{code:java}
int milliVcoresUsed = (int) (cpuUsageTotalCoresPercentage * 1000 * maxVCoresAllottedForContainers
/nodeCpuPercentageForYARN);{code}
 

should be 

 
{code:java}
int milliVcoresUsed = (int) (cpuUsageTotalCoresPercentage * 1000 * maxVCoresAllottedForContainers;
{code}
 

 

I think it need not to divide nodeCpuPercentageForYARN, [~kasha], looks you add this feature,
could you help to have a look ? or could you educate me if I am wrong ?

 

 

 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-dev-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-dev-help@hadoop.apache.org


Mime
View raw message