[
https://issues.apache.org/jira/browse/AMBARI-25278?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Gabor Boros resolved AMBARI-25278.
----------------------------------
Resolution: Fixed
> Fix aggregate metric in Kafka's Grafana dashboard
> -------------------------------------------------
>
> Key: AMBARI-25278
> URL: https://issues.apache.org/jira/browse/AMBARI-25278
> Project: Ambari
> Issue Type: Bug
> Components: ambari-server
> Affects Versions: 2.7.3
> Reporter: Balázs Bence Sári
> Assignee: Balázs Bence Sári
> Priority: Major
> Labels: pull-request-available
> Fix For: 2.7.4
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> The following metrics display aggregate numbers over time on Grafana (they
> are ever increasing). They should show timely rate instead.
> The following metrics are affected:
> *Kafka Home:*
> {code}
> kafka.server.BrokerTopicMetrics.BytesInPerSec.count
> kafka.server.BrokerTopicMetrics.BytesOutPerSec.count
> kafka.server.BrokerTopicMetrics.MessagesInPerSec.count
> kafka.network.RequestMetrics.RequestsPerSec.request.Produce.count
> kafka.network.RequestMetrics.RequestsPerSec.request.FetchConsumer.count
> kafka.controller.ControllerStats.LeaderElectionRateAndTimeMs.count
> kafka.controller.ControllerStats.UncleanLeaderElectionsPerSec.count
> kafka.server.ReplicaManager.IsrShrinksPerSec.count
> kafka.server.ReplicaManager.IsrExpandsPerSec.count
> {code}
> *Kafka Hosts:*
> {code}
> kafka.server.BrokerTopicMetrics.BytesInPerSec.count
> kafka.server.BrokerTopicMetrics.BytesOutPerSec.count
> kafka.server.BrokerTopicMetrics.MessagesInPerSec.count
> kafka.network.RequestMetrics.RequestsPerSec.request.Produce.count
> kafka.network.RequestMetrics.RequestsPerSec.request.FetchConsumer.count
> kafka.server.ReplicaManager.IsrShrinksPerSec.count
> kafka.server.ReplicaManager.IsrExpandsPerSec.count
> {code}
> *Kafka Topics*
> {code}
> kafka.server.BrokerTopicMetrics.BytesInPerSec.topic.*.count
> kafka.server.BrokerTopicMetrics.BytesOutPerSec.topic.*.count
> kafka.server.BrokerTopicMetrics.BytesInPerSec.topic.*.count
> kafka.server.BrokerTopicMetrics.MessagesInPerSec.topic.*.count
> kafka.server.BrokerTopicMetrics.TotalProduceRequestsPerSec.topic.*.count
> {code}
> They are typically needed to be converted from something like this:
> {code}
> "targets": [
> {
> "aggregator": "avg",
> "alias": "ISR Shrinks",
> "app": "kafka_broker",
> "downsampleAggregator": "avg",
> "errors": {},
> "metric": "kafka.server.ReplicaManager.IsrShrinksPerSec.count",
> "precision": "default",
> "refId": "A",
> "transform": "none",
> "transformData": "none",
> "seriesAggregator": "none"
> }
> ],
> {code}
> to this:
> {code}
> "targets": [
> {
> "aggregator": "avg",
> "alias": "ISR Shrinks",
> "app": "kafka_broker",
> "downsampleAggregator": "avg",
> "errors": {},
> "metric": "kafka.server.ReplicaManager.IsrShrinksPerSec.count",
> "precision": "default",
> "refId": "A",
> "transform": "rate",
> "transformData": "none",
> "seriesAggregator": "none"
> }
> ],
> {code}
> with topics metrics, _aggregator_ must be changed to _none_ in addition.
> The Grafana dashboards are stored in the stack definitions.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
