[ 
https://issues.apache.org/jira/browse/HDFS-8672?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15584097#comment-15584097
 ] 

Rui Gao commented on HDFS-8672:
-------------------------------

Hi [~andrew.wang], I agree with you. Separating  recovery-related metrics of EC 
would be very helpful for monitoring. 

Also, I think it's better to keep name consistency among configuration 
parameter, variable name and metric name as mentioned in HDFS-10999 .

> Erasure Coding: Add EC-related Metrics to NN (seperate striped blocks count 
> from UnderReplicatedBlocks count)
> -------------------------------------------------------------------------------------------------------------
>
>                 Key: HDFS-8672
>                 URL: https://issues.apache.org/jira/browse/HDFS-8672
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>            Reporter: Walter Su
>            Assignee: Walter Su
>            Priority: Minor
>
> 1. {{MissingBlocks}} metric is updated in HDFS-8461 so it includes striped 
> blocks.
> 2. {{CorruptBlocks}} metric is updated in HDFS-8619 so it includes striped 
> blocks.
> 3. {{UnderReplicatedBlocks}} and {{PendingReplicationBlocks}} includes 
> striped blocks (HDFS-7912).
> This jira aims to seperate striped blocks count from 
> {{UnderReplicatedBlocks}} count.
> EC file recovery need coding. It's more expensive than block duplication.
> It's necessary to seperate striped blocks count from UnderReplicatedBlocks 
> count. So user can know what's going on.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org

Reply via email to