codelipenghui edited a comment on issue #3059: Behavior of redelivery and ackTimeout in Java client URL: https://github.com/apache/pulsar/issues/3059#issuecomment-442286748 I tried the program in java client version 2.1.0-incubating ``` 2018-11-28T01:34:26.493Z consume 0 2018-11-28T01:34:27.498Z consume 1 2018-11-28T01:34:28.515Z consume 2 2018-11-28T01:34:29.553Z consume 3 2018-11-28T01:34:30.571Z consume 4 2018-11-28T01:34:35.178Z consume 0 2018-11-28T01:34:35.179Z consume 1 2018-11-28T01:34:35.179Z consume 2 2018-11-28T01:34:35.180Z consume 3 2018-11-28T01:34:35.180Z consume 4 2018-11-28T01:34:45.184Z consume 0 2018-11-28T01:34:45.184Z consume 1 2018-11-28T01:34:45.185Z consume 2 2018-11-28T01:34:45.185Z consume 3 2018-11-28T01:34:45.186Z consume 4 ``` @jiazhai looks like caused by the PR #2590 I think primary cause is low granularity time tracking(ack timeout) for messages in a set. should we consider use HashedWheelTimer to instead.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
