[
https://issues.apache.org/jira/browse/HDFS-6881?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14158505#comment-14158505
]
Colin Patrick McCabe commented on HDFS-6881:
--------------------------------------------
I thought about this a little bit more, and I'm not sure my approach here is
the best.
HTrace has this concept of "root spans" and I'm not sure that RPC spans should
be selected to be root spans. RPCs are always in the service of some other
operation, so it might be best to start tracing on the other operation.
To given an example, I think operations like {{DFSClient#setPermission}} make
more sense as root spans than {{ClientProtocol#setPermission}}. After all, if
we're undergoing NameNode failover, or just network congestion, we may have to
make multiple RPC attempts (via ClientProtocol) in order to perform
{{DFSClient#setPermission}}. If we don't have a trace span for
{{DFSClient#setPermission}} operation, we'd never know that these multiple
attempts were done.
So I'm going to close this as WONTFIX and create another change to wrap these
operations in DFSClient.
> The DFSClient should use the sampler to determine whether to initiate trace
> spans when making RPCv9 calls to the NN and DN
> --------------------------------------------------------------------------------------------------------------------------
>
> Key: HDFS-6881
> URL: https://issues.apache.org/jira/browse/HDFS-6881
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: hdfs-client
> Affects Versions: 2.6.0
> Reporter: Masatake Iwasaki
> Assignee: Colin Patrick McCabe
> Attachments: HDFS-6881.002.patch, HDFS-6881.003.patch
>
>
> The DFSClient should use the configred HTrace sampler to determine whether to
> initiate trace spans when making RPCv9 calls to the NN and DN.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
