[
https://issues.apache.org/jira/browse/FLINK-4563?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15529746#comment-15529746
]
Anton Mushin commented on FLINK-4563:
-------------------------------------
I am understand.
I planning to fix this issue after merge
https://github.com/apache/flink/pull/2517 or when PR will be accept.
Now in AbstractMetricGroup#getMetricIdentifier() and for all calls
getMetricIdentifier method scopeString once initialized.
{code:java}
public String getMetricIdentifier(String metricName, CharacterFilter filter) {
if (scopeString == null) {
if (filter != null) {
scopeString = ScopeFormat.concat(filter,
registry.getDelimiter(), scopeComponents);
} else {
scopeString =
ScopeFormat.concat(registry.getDelimiter(), scopeComponents);
}
}
if (filter != null) {
return scopeString + registry.getDelimiter() +
filter.filterCharacters(metricName);
} else {
return scopeString + registry.getDelimiter() +
metricName;
}
}
{code}
Do I need save this "caching" of scopeString for different filters?
It is variable using in 3 places class:
1 - in getMetricIdentifier() method
2 - in addGroup() method
3 - in addMetric() method
For 2 and 3 cases 'scopeString' using for logging
{code:title=Example}
LOG.warn("Name collision: Adding a metric with the same name as a metric
subgroup: '" +
name + "'.
Metric might not get properly reported. (" + scopeString + ')');
{code}
As solution this issue I think what I can remove this variable from global
scope and initialize it every time when call getMetric Identifier. But it is
simple and most likely untrue solution.
> [metrics] scope caching not adjusted for multiple reporters
> -----------------------------------------------------------
>
> Key: FLINK-4563
> URL: https://issues.apache.org/jira/browse/FLINK-4563
> Project: Flink
> Issue Type: Bug
> Components: Metrics
> Affects Versions: 1.1.0
> Reporter: Chesnay Schepler
> Assignee: Anton Mushin
>
> Every metric group contains a scope string, representing what entities
> (job/task/etc.) a given metric belongs to, which is calculated on demand.
> Before this string is cached a CharacterFilter is applied to it, which is
> provided by the callee, usually a reporter. This was done since different
> reporters have different requirements in regards to valid characters. The
> filtered string is cached so that we don't have to refilter the string every
> time.
> This all works fine with a single reporter; with multiple however it is
> completely broken as only the first filter is ever applied.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
