[
https://issues.apache.org/jira/browse/AMBARI-17843?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15388546#comment-15388546
]
Siddharth Wagle commented on AMBARI-17843:
------------------------------------------
+ LGTM, does the metadata population work for APP metrics and APP_HOST metrics?
> App data aggregated for hosted apps is being calcualted for all apps, not
> just configured ones
> ----------------------------------------------------------------------------------------------
>
> Key: AMBARI-17843
> URL: https://issues.apache.org/jira/browse/AMBARI-17843
> Project: Ambari
> Issue Type: Bug
> Components: ambari-metrics
> Affects Versions: 2.4.0
> Reporter: Aravindan Vijayan
> Assignee: Aravindan Vijayan
> Fix For: 2.4.0
>
> Attachments: AMBARI-17843.patch
>
>
> timeline.metrics.service.cluster.aggregator.appIds defines what apps to
> produce host metrics for:
> {quote}List of application ids to use for aggregating host level metrics for
> an application. Example: bytes_read across Yarn Nodemanagers. {quote}
> Right now we are aggregating for all.
> Additionally, metadata does not expose these additional metrics
> {code}
> 0: jdbc:phoenix:localhost:61181:/ams-hbase-un> select distinct(APP_ID) from
> METRIC_AGGREGATE_MINUTE WHERE METRIC_NAME = 'cpu_user';
> +------------+
> | APP_ID |
> +------------+
> | HOST |
> | ams-hbase |
> | amssmoketestfake |
> | applicationhistoryserver |
> | datanode |
> | hivemetastore |
> | hiveserver2 |
> | jobhistoryserver |
> | namenode |
> | nimbus |
> | nodemanager |
> | resourcemanager |
> +------------+
> 12 rows selected (0.117 seconds)
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
