[ 
https://issues.apache.org/jira/browse/HDFS-5055?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13725877#comment-13725877
 ] 

Allen Wittenauer commented on HDFS-5055:
----------------------------------------

On our machines, we create a VIP using a BGP loopback address.  We bind Hadoop 
services to this VIP.  In the case of the NN trying to reconnect back to the 
2NN during edits merge, it continues to use the 2NN's hostname address:

In hdfs-site.xml:

{code}
  <property>
    <name>dfs.namenode.secondary.http-address</name>
    <value>eat1-fiasconn02.grid.linkedin.com:50090</value>
  </property>
{code}

On the 2nn itself:

{code}
[awittena@eat1-hcl4054 ~]$ netstat -a | grep 50090
tcp        0      0 eat1-fiasconn02.grid.:50090 *:*                         
LISTEN      
[awittena@eat1-hcl4054 ~]$ 
{code}

yet in the HDFS NN log...

{code}
2013-07-31 23:53:30,916 INFO 
org.apache.hadoop.hdfs.server.namenode.TransferFsImage: Opening connection to 
http://eat1-hcl4054.grid.linkedin.com:50090/getimage?getimage=1&txid=180564&storageInfo=-47:2073800469:1375222008066:CID-1289fa67-537a-4c27-9bba-94770b94644d
2013-07-31 23:53:30,917 ERROR org.apache.hadoop.security.UserGroupInformation: 
PriviledgedActionException 
as:hdfs/eat1-fiasconn01.grid.linkedin....@grid.linkedin.com (auth:KERBEROS) 
cause:java.net.ConnectException: Connection refused
2013-07-31 23:53:30,919 WARN org.mortbay.log: /getimage: java.io.IOException: 
GetImage failed. java.net.ConnectException: Connection refused
{code}

This is a regression from 1.x behavior.
                
> nn<->2nn ignore dfs.namenode.secondary.http-address
> ---------------------------------------------------
>
>                 Key: HDFS-5055
>                 URL: https://issues.apache.org/jira/browse/HDFS-5055
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: namenode
>    Affects Versions: 2.1.0-beta
>            Reporter: Allen Wittenauer
>            Priority: Blocker
>
> The primary namenode attempts to connect back to (incoming hostname):port 
> regardless of how dfs.namenode.secondary.http-address is configured.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to