[
https://issues.apache.org/jira/browse/GIRAPH-985?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14288030#comment-14288030
]
Hudson commented on GIRAPH-985:
-------------------------------
FAILURE: Integrated in Giraph-trunk-Commit #1521 (See
[https://builds.apache.org/job/Giraph-trunk-Commit/1521/])
GIRAPH-985 addition: Making sum counter come first (majakabiljo:
http://git-wip-us.apache.org/repos/asf?p=giraph.git&a=commit&h=a6f2a4fc929d01073058578175981b0c1c616ba0)
*
giraph-core/src/main/java/org/apache/giraph/metrics/SuperstepMetricsRegistry.java
> Add more metrics
> ----------------
>
> Key: GIRAPH-985
> URL: https://issues.apache.org/jira/browse/GIRAPH-985
> Project: Giraph
> Issue Type: Improvement
> Reporter: Maja Kabiljo
> Assignee: Maja Kabiljo
>
> When limit number of open requests is on, significant part of computation can
> be spent in just waiting. So adding metrics for total amount of time compute
> threads spent in waiting on open requests, and histogram with compute times
> per partition. These should help detecting why some job is slower than
> expected (is it from messaging or compute).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
