[
https://issues.apache.org/jira/browse/AMBARI-10194?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14379079#comment-14379079
]
Hudson commented on AMBARI-10194:
---------------------------------
ABORTED: Integrated in Ambari-trunk-Commit #2104 (See
[https://builds.apache.org/job/Ambari-trunk-Commit/2104/])
AMBARI-10194. Better Client Side Validation Of Alert Definition Threshold Data
(alexantonenko) (hiveww:
http://git-wip-us.apache.org/repos/asf?p=ambari.git&a=commit&h=c42db2ebe37b4658c8b5a52024520c9ea7875d93)
* ambari-web/app/models/alert_config.js
* ambari-web/test/models/alert_config_test.js
* ambari-web/app/utils/number_utils.js
> Better Client Side Validation Of Alert Definition Threshold Data
> ----------------------------------------------------------------
>
> Key: AMBARI-10194
> URL: https://issues.apache.org/jira/browse/AMBARI-10194
> Project: Ambari
> Issue Type: Task
> Components: ambari-web
> Affects Versions: 2.0.0
> Reporter: Antonenko Alexander
> Assignee: Antonenko Alexander
> Fix For: 2.1.0
>
> Attachments: AMBARI-10194.patch
>
>
> When editing the alert definition thresholds, the web client currently allows
> precision that is far beyond what makes sense for the threshold values. For
> example, if you go to edit the thresholds for the seconds that a port alert
> will use, you can set the value up to the micro second: 5.0000001. Values
> beyond this are actually not considered changes and not sent to the backend
> even though the web client indicates that a change was saved.
> In general, there should be smarter precision for the threshold values:
> - AGGREGATE: whole percent values (80%, not 80.0%)
> - PORT: 1/10th of a second: (1.5s, 5.0s)
> - METRIC: Current values are whole numbers and I don't really see the need to
> make this any more precise. Although the nature of metrics dictates that any
> value could be compared, including values less than 1. I'd say allow the same
> as for PORT; down to the 1/10th precision.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
