[
https://issues.apache.org/jira/browse/HDFS-14406?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16812511#comment-16812511
]
Erik Krogen commented on HDFS-14406:
------------------------------------
I'd like to echo [~daryn]'s sentiment that this probably won't provide much
useful information for HDFS given the lock behavior. You can see more
discussion of this in HDFS-14403, a recently updated proposal to extend the
FairCallQueue to consider cost of an operation (in the case of HDFS, this is
considered to be lock hold time). The design document has some extended
discussion of why looking at the overall processing time is insufficient for
HDFS. In that patch, we've surfaced information about the lock hold time itself
up to the RPC layer; maybe it would make more sense to surface that value
rather than the processing time?
[~daryn], I'd love to learn more about your patch for excluding lock wait time
from processing time.
> Add per user RPC Processing time
> --------------------------------
>
> Key: HDFS-14406
> URL: https://issues.apache.org/jira/browse/HDFS-14406
> Project: Hadoop HDFS
> Issue Type: Improvement
> Affects Versions: 3.2.0
> Reporter: Xue Liu
> Assignee: Xue Liu
> Priority: Minor
> Fix For: 3.2.0
>
> Attachments: HDFS-14406.001.patch
>
>
> For a shared cluster we would want to separate users' resources, as well as
> having our metrics reflecting on the usage, latency, etc, for each user.
> This JIRA aims to add per user RPC processing time metrics and expose it via
> JMX.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
