[
https://issues.apache.org/jira/browse/HDFS-9579?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15202289#comment-15202289
]
Sangjin Lee commented on HDFS-9579:
-----------------------------------
The latest patch looks good, but {{ReplicaAccessor}} adds a new abstract
method. Since it's declared as public/stable, this does raise the question of
compatibility. I'm not too familiar with that class and it's not clear whether
it was meant to be implemented only by Hadoop code or can be implemented by
users/libraries. [~cmccabe], thoughts on this?
> Provide bytes-read-by-network-distance metrics at FileSystem.Statistics level
> -----------------------------------------------------------------------------
>
> Key: HDFS-9579
> URL: https://issues.apache.org/jira/browse/HDFS-9579
> Project: Hadoop HDFS
> Issue Type: Improvement
> Reporter: Ming Ma
> Assignee: Ming Ma
> Attachments: HDFS-9579-2.patch, HDFS-9579-3.patch, HDFS-9579-4.patch,
> HDFS-9579-5.patch, HDFS-9579-6.patch, HDFS-9579-7.patch, HDFS-9579-8.patch,
> HDFS-9579-9.patch, HDFS-9579.patch, MR job counters.png
>
>
> For cross DC distcp or other applications, it becomes useful to have insight
> as to the traffic volume for each network distance to distinguish cross-DC
> traffic, local-DC-remote-rack, etc.
> FileSystem's existing {{bytesRead}} metrics tracks all the bytes read. To
> provide additional metrics for each network distance, we can add additional
> metrics to FileSystem level and have {{DFSInputStream}} update the value
> based on the network distance between client and the datanode.
> {{DFSClient}} will resolve client machine's network location as part of its
> initialization. It doesn't need to resolve datanode's network location for
> each read as {{DatanodeInfo}} already has the info.
> There are existing HDFS specific metrics such as {{ReadStatistics}} and
> {{DFSHedgedReadMetrics}}. But these metrics are only accessible via
> {{DFSClient}} or {{DFSInputStream}}. Not something that application framework
> such as MR and Tez can get to. That is the benefit of storing these new
> metrics in FileSystem.Statistics.
> This jira only includes metrics generation by HDFS. The consumption of these
> metrics at MR and Tez will be tracked by separated jiras.
> We can add similar metrics for HDFS write scenario later if it is necessary.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
