[
https://issues.apache.org/jira/browse/HADOOP-15611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16552253#comment-16552253
]
Yiqun Lin commented on HADOOP-15611:
------------------------------------
Thanks [~jianliang.wu] for attaching the patch! Some minor comments from me:
* Use DEBUG log level to print for {{Start decay current counts.}}, and update
{{Start}} to {{Start to}}.
* {{"The decayedCount for this user "+entry.getKey().toString()+" is zero"}}
can be updated as {{"The decayed count for the user {} is zero and being
cleaned.",
* We need to add a space before {{totalRawCallCount}} in {{"After decaying the
stored counts, totalDecayedCount: {}," + "totalRawCallCount: {}.}}
* Remove following unrelated change:
{code}
if (totalCallSnapshot > 0) {
proportion = (double) occurrences / totalCallSnapshot;
}
-
// Start with low priority levels, since they will be most common
{code}
* {{"Moving to next queue, next queue index : {}," + "next queue weight : {},
the requestsLeft : {}."}} can be updated as {{"Moving to next queue from queue
index {} to index {}, number of requests left for current queue: {}."
BTW, you should also format these lines.
> Improve log in FairCallQueue
> ----------------------------
>
> Key: HADOOP-15611
> URL: https://issues.apache.org/jira/browse/HADOOP-15611
> Project: Hadoop Common
> Issue Type: Improvement
> Affects Versions: 3.1.0
> Reporter: Ryan Wu
> Priority: Minor
> Attachments: HADOOP-15611.001.patch
>
>
> In the usage of the FairCallQueue, we find there missing some Key log. Only a
> few logs are printed, it makes us hard to learn and debug this feature.
> At least, following places can print more logs.
> * DecayRpcScheduler#decayCurrentCounts
> * WeightedRoundRobinMultiplexer#moveToNextQueue
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
