[
https://issues.apache.org/jira/browse/HDFS-14940?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16977706#comment-16977706
]
Ayush Saxena commented on HDFS-14940:
-------------------------------------
{quote}Why to send one extra RPC to server if one common place available to
handle this in client side (FS API).
{quote}
I too, though this only could be one reason, but we restrict this functionality
from the Client, but Namenode still allows this, somebody directly connecting
through a Client, can still go ahead with it. Shouldn't we have it at Namenode
side too, as you said before and we can have it at the client side too saving
the RPC.
> HDFS Balancer : Do not allow to set balancer maximum network bandwidth more
> than 1TB
> ------------------------------------------------------------------------------------
>
> Key: HDFS-14940
> URL: https://issues.apache.org/jira/browse/HDFS-14940
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: balancer & mover
> Affects Versions: 3.1.1
> Environment: 3 Node HA Setup
> Reporter: Souryakanta Dwivedy
> Priority: Minor
> Attachments: BalancerBW.PNG, HDFS-14940.001.patch,
> HDFS-14940.002.patch, HDFS-14940.003.patch
>
>
> HDFS Balancer : getBalancerBandwidth displaying wrong values for the maximum
> network bandwidth used by the datanode
> while network bandwidth set with values as 1048576000g/1048p/1e
> Steps :-
> * Set balancer bandwith with command setBalancerBandwidth and vlaues as
> [1048576000g/1048p/1e]
> * - Check bandwidth used by the datanode during HDFS block balancing with
> command :hdfs dfsadmin -getBalancerBandwidth " check it will display some
> different values not the same value as set
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
