[ 
https://issues.apache.org/jira/browse/HADOOP-18691?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17710143#comment-17710143
 ] 

ASF GitHub Bot commented on HADOOP-18691:
-----------------------------------------

cxzl25 opened a new pull request, #5542:
URL: https://github.com/apache/hadoop/pull/5542

   ### Description of PR
   When `Client.Connection#setupConnection` encounters an `IOException`, it 
will try to update the server address. 
([HADOOP-18365](https://issues.apache.org/jira/browse/HADOOP-18365))
   
   When the address is re-parsed, it may be an unresolved address 
(`UnknownHostException`), which causes `Client.Connection#setupConnection` to 
fail to reconnect.
   
   ```java
   while (true) {
     try {
       if (server.isUnresolved()) { 
   ```
   
   Especially when DN is connected to NN, `BPServiceActor#bpNamenode` is only 
initialized once, which causes DN to never connect to NN before restarting.
   
   ### How was this patch tested?
   
   
   ### For code changes:
   
   - [ ] Does the title or this PR starts with the corresponding JIRA issue id 
(e.g. 'HADOOP-17799. Your PR title ...')?
   - [ ] Object storage: have the integration tests been executed and the 
endpoint declared according to the connector-specific documentation?
   - [ ] If adding new dependencies to the code, are these dependencies 
licensed in a way that is compatible for inclusion under [ASF 
2.0](http://www.apache.org/legal/resolved.html#category-a)?
   - [ ] If applicable, have you updated the `LICENSE`, `LICENSE-binary`, 
`NOTICE-binary` files?
   
   




> Add a CallerContext getter on the Schedulable interface
> -------------------------------------------------------
>
>                 Key: HADOOP-18691
>                 URL: https://issues.apache.org/jira/browse/HADOOP-18691
>             Project: Hadoop Common
>          Issue Type: Improvement
>            Reporter: Christos Bisias
>            Priority: Major
>              Labels: pull-request-available
>
> We would like to add a default *{color:#00875a}CallerContext{color}* getter 
> on the *{color:#00875a}Schedulable{color}* interface
> {code:java}
> default public CallerContext getCallerContext() {
>   return null;  
> } {code}
> and then override it on the 
> *{color:#00875a}i{color}{color:#00875a}{*}pc/{*}Server.Call{color}* class
> {code:java}
> @Override
> public CallerContext getCallerContext() {  
>   return this.callerContext;
> } {code}
> to expose the already existing *{color:#00875a}callerContext{color}* field.
>  
> This change will help us access the *{color:#00875a}CallerContext{color}* on 
> an Apache Ozone *{color:#00875a}IdentityProvider{color}* implementation.
> On Ozone side the *{color:#00875a}FairCallQueue{color}* doesn't work with the 
> Ozone S3G, because all users are masked under a special S3G user and there is 
> no impersonation. Therefore, the FCQ reads only 1 user and becomes 
> ineffective. We can use the *{color:#00875a}CallerContext{color}* field to 
> store the current user and access it on the Ozone 
> {*}{color:#00875a}IdentityProvider{color}{*}.
>  
> This is a presentation with the proposed approach.
> [https://docs.google.com/presentation/d/1iChpCz_qf-LXiPyvotpOGiZ31yEUyxAdU4RhWMKo0c0/edit#slide=id.p]



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

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

Reply via email to