[ 
https://issues.apache.org/jira/browse/HDFS-13522?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17568244#comment-17568244
 ] 

ZanderXu commented on HDFS-13522:
---------------------------------

Thanks [~simbadzina] for your design doc. 
We should need a more detailed design doc to explain that why and how RBF 
supports Observer-Read? 

I have a few questions about it, you can refer to when you are designing.
# What's the overall forwarding process?
# Does the client need to carry the StateID of all NSs? Is this function needed 
in the Hadoop Ecosystem? It's very expensive.
# How does the client enable or disable this feature? Latency-sensitive clients 
may choose to disable this feature.
# How does RBF monitor and store the status of OBserver NN?
# How does RBF forward the read requests to OBserver NN? such as how to msync? 
how to obtain namenode list?
# Does RBF need to consider the scenario of OBserver NN HA switching?
# How does RBF control and store the StateId from NameNodes?

> Huawei has one, Uber has a different one and now this third one..

We'd better discuss the overall implementation plan first, and then start 
development and code review after reaching an agreement, how? and we can create 
several subTasks to do this step by step, and it will be very helpful for code 
review and code merging.

[~simbadzina]Thanks for your works again. This is a very helpful feature for 
end users, and I really appreciate it and hope you can keep pushing it. 





> RBF: Support observer node from Router-Based Federation
> -------------------------------------------------------
>
>                 Key: HDFS-13522
>                 URL: https://issues.apache.org/jira/browse/HDFS-13522
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: federation, namenode
>            Reporter: Erik Krogen
>            Assignee: Simbarashe Dzinamarira
>            Priority: Major
>              Labels: pull-request-available
>         Attachments: HDFS-13522.001.patch, HDFS-13522.002.patch, 
> HDFS-13522_WIP.patch, RBF_ Observer support.pdf, Router+Observer RPC 
> clogging.png, ShortTerm-Routers+Observer.png, 
> observer_reads_in_rbf_proposal_simbadzina_v1.pdf
>
>          Time Spent: 19h
>  Remaining Estimate: 0h
>
> Changes will need to occur to the router to support the new observer node.
> One such change will be to make the router understand the observer state, 
> e.g. {{FederationNamenodeServiceState}}.



--
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