[
https://issues.apache.org/jira/browse/EAGLE-438?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15414453#comment-15414453
]
ASF GitHub Bot commented on EAGLE-438:
--------------------------------------
Github user RalphSu commented on a diff in the pull request:
https://github.com/apache/incubator-eagle/pull/316#discussion_r74164639
--- Diff: eagle-external/hadoop_jmx_collector/config-sample.json ---
@@ -12,18 +12,35 @@
"https":false
}
},
- "input": {
- "component": "namenode",
- "port": "50070",
- "https": false
- },
+ "inputs": [
+ {
+ "component": "namenode",
+ "host": "127.0.0.1",
+ "port": "50070",
+ "https": false,
+ "kafka_topic": "nn_jmx_metric_sandbox"
+ },
+ {
+ "component": "resourcemanager",
+ "host": "127.0.0.1",
+ "port": "8088",
+ "https": false,
+ "kafka_topic": "rm_jmx_metric_sandbox"
+ },
+ {
+ "component": "datanode",
+ "host": "127.0.0.1",
--- End diff --
let's use place holder like what we do in PR312, use
server.eagle.apache.org. So that developer could simply use /etc/resolv.conf
when there is different endpoint he want to change (etc, hdp sandbox)
> Multiple Inputs for Hadoop JMX Collector Python Script
> ------------------------------------------------------
>
> Key: EAGLE-438
> URL: https://issues.apache.org/jira/browse/EAGLE-438
> Project: Eagle
> Issue Type: New Feature
> Reporter: Peter Kim
> Priority: Trivial
> Labels: features
>
> It would be very useful for the hadoop jmx collector to be able to collect
> jmx metrics for multiple components at once. So here, I wish to extend the
> default python script that collects Hadoop JMX Metrics to support multiple
> inputs as oppose to only one. So, with this, one can collect all jmx metrics
> for multiple components such as namenode, datanode, resource manager,
> hmaster, etc. all at once.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
