[GitHub] [pulsar] codelipenghui commented on a change in pull request #10168: fix 10169 Delayed messages may be executed several milliseconds in advance

[GitBox <gi...@apache.org>]

codelipenghui commented on a change in pull request #10168:
URL: https://github.com/apache/pulsar/pull/10168#discussion_r611187892



##########
File path: pulsar-broker-common/src/main/java/org/apache/pulsar/broker/ServiceConfiguration.java
##########
@@ -238,6 +238,9 @@
             + " affecting the accuracy of the delivery time compared to the scheduled time. Default is 1 second.")
     private long delayedDeliveryTickTimeMillis = 1000;
 
+    @FieldContext(category = CATEGORY_SERVER, doc = "HashedWheelTimer duration tick time for when retrying on delayed delivery")
+    private long delayedDeliveryTickDurationMillis = 1000;

Review comment:
       @baomingyu I don't think we need this config there. For most cases and the current implementation of Pulsar delay message, accuracy in the millisecond level is the expected behavior, If the default value 1000 is not work for you, I think you can decrease it to 5ms or lower. Introduce 2 configurations to control the tick time of the delayed message tracker also will introduce complexity to users to tune the delayed message behavior.

##########
File path: pulsar-broker-common/src/main/java/org/apache/pulsar/broker/ServiceConfiguration.java
##########
@@ -1510,6 +1513,8 @@
     private String managedLedgerDataReadPriority = OffloadPolicies.OffloadedReadPriority.TIERED_STORAGE_FIRST
             .getValue();
 
+    private long stickConsumerBackOffDelayTimeMs = 500;

Review comment:
       Seems not related to this PR.




-- 
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.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org