gyfora commented on code in PR #686:
URL:
https://github.com/apache/flink-kubernetes-operator/pull/686#discussion_r2286971695
##########
flink-autoscaler/src/main/java/org/apache/flink/autoscaler/metrics/ScalingMetrics.java:
##########
@@ -78,6 +88,45 @@ public static void computeDataRateMetrics(
}
}
+ private static Optional<Double> getObservedTpr(
+ Map<FlinkMetric, AggregatedMetric> flinkMetrics,
+ Map<ScalingMetric, Double> scalingMetrics,
+ double numRecordsInPerSecond,
+ Configuration conf) {
+
+ // If there are no incoming records we return infinity to allow scale
down
+ if (numRecordsInPerSecond == 0) {
+ return Optional.of(Double.POSITIVE_INFINITY);
+ }
+
+ // We only measure observed tpr when we are catching up, that is when
the lag is beyond the
+ // configured observe threshold
+ boolean catchingUp =
+ scalingMetrics.getOrDefault(ScalingMetric.LAG, 0.)
+ >=
conf.get(AutoScalerOptions.OBSERVE_TRUE_PROCESSING_RATE_LAG_THRESHOLD)
+ .toSeconds()
+ * numRecordsInPerSecond;
+ if (!catchingUp) {
+ return Optional.empty();
+ }
+
+ double observedTpr =
+ computeObservedTprWithBackpressure(
+ numRecordsInPerSecond,
+
flinkMetrics.get(FlinkMetric.BACKPRESSURE_TIME_PER_SEC).getAvg());
+
+ return Double.isNaN(observedTpr) ? Optional.empty() :
Optional.of(observedTpr);
+ }
+
+ public static double computeObservedTprWithBackpressure(
Review Comment:
observed TPR has not much to do with catch up time. It was introduced
specifically to compute processing capacity (TPR) correctly for sources in
cases when the busy time reported for Flink is completely off. You can read
more about this in the description
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
