(kafka) branch trunk updated: MINOR : Corrected the stdout parsing in ShareConsumerPerformanceService (#21473)

[schofielaj]

This is an automated email from the ASF dual-hosted git repository.

schofielaj pushed a commit to branch trunk
in repository https://gitbox.apache.org/repos/asf/kafka.git


The following commit(s) were added to refs/heads/trunk by this push:
     new 9e1f91d6264 MINOR : Corrected the stdout parsing in 
ShareConsumerPerformanceService (#21473)
9e1f91d6264 is described below

commit 9e1f91d6264dfd6974485b967c3129539d1803dc
Author: Chirag Wadhwa <cwad...@confluent.io>
AuthorDate: Sat Feb 14 03:22:34 2026 +0530

    MINOR : Corrected the stdout parsing in ShareConsumerPerformanceService 
(#21473)
    
    The ShareConsumerPerformanceService runs the
    `kafka-share-consumer-perf-test.sh` script and parses its stdout to
    further report the performance numbers. The script's output format is as
    follows ->
    
    `start.time, end.time, data.consumed.in.MB, MB.sec, nMsg.sec,
    data.consumed.in.nMsg, fetch.time.ms`
    
    One of the fields it tried to read is `records_per_sec`. As per the
    current code, it parses the wrong value (instead of reading the 5th
    value (nMsg.sec), it reads the 6th one (data.consumed.in.nMsg)). This PR
    corrects that.
    
    Reviewers: Andrew Schofield <aschofi...@confluent.io>
---
 tests/kafkatest/services/performance/share_consumer_performance.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/tests/kafkatest/services/performance/share_consumer_performance.py 
b/tests/kafkatest/services/performance/share_consumer_performance.py
index 432d1e1da81..1cc79bc18ae 100644
--- a/tests/kafkatest/services/performance/share_consumer_performance.py
+++ b/tests/kafkatest/services/performance/share_consumer_performance.py
@@ -131,10 +131,13 @@ class ShareConsumerPerformanceService(PerformanceService):
             last = line
 
         # Parse and save the last line's information
+        # The script kafka-share-consumer-perf-test.sh first prints the header 
line in the following format:
+        #   start.time, end.time, data.consumed.in.MB, MB.sec, nMsg.sec, 
data.consumed.in.nMsg, fetch.time.ms
+        # The corresponding results are then printed in the same order, so we 
can parse the last line for the results.
         if last is not None:
             parts = last.split(',')
             self.results[idx-1] = {
                 'total_mb': float(parts[2]),
                 'mbps': float(parts[3]),
-                'records_per_sec': float(parts[5]),
+                'records_per_sec': float(parts[4]),
             }