Ewen Cheslack-Postava created KAFKA-4526:
--------------------------------------------
Summary: Transient failure in
ThrottlingTest.test_throttled_reassignment
Key: KAFKA-4526
URL: https://issues.apache.org/jira/browse/KAFKA-4526
Project: Kafka
Issue Type: Bug
Reporter: Ewen Cheslack-Postava
Assignee: Jason Gustafson
Fix For: 0.10.2.0
This test is seeing transient failures sometimes
{quote}
Module: kafkatest.tests.core.throttling_test
Class: ThrottlingTest
Method: test_throttled_reassignment
Arguments:
{
"bounce_brokers": false
}
{quote}
This happens with both bounce_brokers = true and false. Fails with
{quote}
AssertionError: 1646 acked message did not make it to the Consumer. They are:
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19...plus
1626 more. Total Acked: 174799, Total Consumed: 173153. We validated that the
first 1000 of these missing messages correctly made it into Kafka's data files.
This suggests they were lost on their way to the consumer.
{quote}
See
http://confluent-kafka-system-test-results.s3-us-west-2.amazonaws.com/2016-12-12--001.1481535295--apache--trunk--62e043a/report.html
for an example.
Note that there are a number of similar bug reports for different tests:
https://issues.apache.org/jira/issues/?jql=text%20~%20%22acked%20message%20did%20not%20make%20it%20to%20the%20Consumer%22%20and%20project%20%3D%20Kafka
I am wondering if we have a wrong ack setting somewhere that we should be
specifying as acks=all but is only defaulting to 0?
It also seems interesting that the missing messages in these recent failures
seem to always start at 0...
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
