You are viewing a plain text version of this content. The canonical link for it is here.

Posted to common-dev@hadoop.apache.org by "Jeper (JIRA)" <ji...@apache.org> on 2018/09/06 09:23:00 UTC

[jira] [Created] (HADOOP-15724) Some rack awareness question

Jeper created HADOOP-15724:
------------------------------

             Summary: Some rack awareness question
                 Key: HADOOP-15724
                 URL: https://issues.apache.org/jira/browse/HADOOP-15724
             Project: Hadoop Common
          Issue Type: Test
         Environment: hadoop-2.5.2
            Reporter: Jeper
         Attachments: image-2018-09-06-17-12-02-610.png

When I use the rack awareness, I stop all the machines in a rack, and then write the data and report the following error，I don't think this is a reasonable design
!image-2018-09-06-17-12-02-610.png!

I think the following design is more reasonable:
 can write, but only one replica, or can write, several copies on the same rack

I have some questions
 1. What's the difference when writing data, between rack awareness and no rack
 2. Join a rack awareness when a running cluster joins, will the old data be distributed to both racks?
Forgive me, I am not very familiar with java programming, can't understand the source code.



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

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