[ 
https://issues.apache.org/jira/browse/HDFS-9756?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16825834#comment-16825834
 ] 

bianqi commented on HDFS-9756:
------------------------------

Why is there no one paying attention to this problem?

> Hard-Code value in DataTransferThrottler
> ----------------------------------------
>
>                 Key: HDFS-9756
>                 URL: https://issues.apache.org/jira/browse/HDFS-9756
>             Project: Hadoop HDFS
>          Issue Type: Bug
>    Affects Versions: 2.7.1
>            Reporter: Yiqun Lin
>            Assignee: Yiqun Lin
>            Priority: Major
>         Attachments: HDFS-9756.001.patch, HDFS-9756.002.patch
>
>
> In DataTransferThrottler, the period time is hard-code for 500 ms. Even 
> though it has other construction method, 
> {code}
> /**
>    * Constructor
>    * @param period in milliseconds. Bandwidth is enforced over this
>    *        period.
>    * @param bandwidthPerSec bandwidth allowed in bytes per second. 
>    */
>   public DataTransferThrottler(long period, long bandwidthPerSec) {
>     this.curPeriodStart = monotonicNow();
>     this.period = period;
>     this.curReserve = this.bytesPerPeriod = bandwidthPerSec*period/1000;
>     this.periodExtension = period*3;
>   }
> {code}
> but it was only invoked by this method 
> {code}
> public DataTransferThrottler(long bandwidthPerSec) {
>     this(500, bandwidthPerSec);  // by default throttling period is 500ms 
> }
> {code}
> So the period is a hard-code. This value can also influence the 
> data-transfering. If period time value set small, the number of times to 
> waitting for next period will be increased and the total waitting-time also  
> be increased. So the average bandwidth will be decreased.



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

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to