[ 
https://issues.apache.org/jira/browse/KNOX-1440?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16629308#comment-16629308
 ] 

Phil Zampino edited comment on KNOX-1440 at 9/26/18 8:37 PM:
-------------------------------------------------------------

The HA dispatch class name for the datanode route can be explicitly overridden 
in the service definition.
{code:java}
<route path="/webhdfs/data/v1/**?**">
  <rewrite apply="WEBHDFS/webhdfs/inbound/datanode" to="request.url"/>
  <dispatch contributor-name="http-client" 
ha-classname="org.apache.knox.gateway.dispatch.DefaultDispatch" />
</route>
{code}
 


was (Author: pzampino):
As a work-around, the HA dispatch class name can be explicitly overridden in 
the service definition.

 
{code:java}
<route path="/webhdfs/data/v1/**?**">
  <rewrite apply="WEBHDFS/webhdfs/inbound/datanode" to="request.url"/>
  <dispatch contributor-name="http-client" 
ha-classname="org.apache.knox.gateway.dispatch.DefaultDispatch" />
</route>
{code}
 

> WEBHDFS: retry should not be attempted for datanode write failures
> ------------------------------------------------------------------
>
>                 Key: KNOX-1440
>                 URL: https://issues.apache.org/jira/browse/KNOX-1440
>             Project: Apache Knox
>          Issue Type: Improvement
>          Components: Server
>    Affects Versions: 1.1.0
>            Reporter: Phil Zampino
>            Assignee: Phil Zampino
>            Priority: Major
>             Fix For: 1.2.0
>
>
> For WEBHDFS invocations through Knox, if a *datanode* write fails with an 
> IOException, the currently configured HA dispatch (DefaultHaDispatch) 
> attempts the write the configured number of times.
> HDFS dev says that it would be appropriate to more simply respond with the 
> error, deferring the decision to retry to the client.
> The best resolution for this is to have the datanode interactions use the 
> DefaultDispatch (rather than the DefaultHaDispatch) even when HA is enabled, 
> such that no retry/failover will be attempted for those interactions.
> (FYI: the datanode dispatch is configured in the WEBHDFS service definition)



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

Reply via email to