[
https://issues.apache.org/jira/browse/FLINK-7502?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16145154#comment-16145154
]
ASF GitHub Bot commented on FLINK-7502:
---------------------------------------
Github user zentol commented on a diff in the pull request:
https://github.com/apache/flink/pull/4586#discussion_r135764533
--- Diff:
flink-metrics/flink-metrics-prometheus/src/main/java/org/apache/flink/metrics/prometheus/PrometheusReporter.java
---
@@ -130,7 +139,11 @@ public void notifyOfAddedMetric(final Metric metric,
final String metricName, fi
metric.getClass().getName());
return;
}
- collector.register();
+ try {
+ collector.register();
+ } catch (Exception e) {
+ LOG.warn("There was a problem registering metric {}:
{}", metricName, e);
--- End diff --
We usually don't include placeholders for exceptions (because they are
added implicitly), i.e. it should be ```LOG.warn("There was a problem
registering metric {}.", metricName, e);```
> PrometheusReporter improvements
> -------------------------------
>
> Key: FLINK-7502
> URL: https://issues.apache.org/jira/browse/FLINK-7502
> Project: Flink
> Issue Type: Improvement
> Components: Metrics
> Affects Versions: 1.4.0
> Reporter: Maximilian Bode
> Assignee: Maximilian Bode
> Priority: Minor
>
> * do not throw exceptions on metrics being registered for second time
> * allow port ranges for setups where multiple reporters are on same host
> (e.g. one TaskManager and one JobManager)
> * do not use nanohttpd anymore, there is now a minimal http server included
> in [Prometheus JVM client|https://github.com/prometheus/client_java]
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
