[GitHub] [kafka] lucasbru commented on pull request #13876: KAFKA-10733: Clean up producer exceptions

2023-07-11 Thread via GitHub
lucasbru commented on PR #13876: URL: https://github.com/apache/kafka/pull/13876#issuecomment-1630247754 @jolshan -- I will discuss the remaining scope with @guozhangwang now that he's back. Can we merge this restricted PR? -- This is an automated message from the Apache Git Service. To

[GitHub] [kafka] lucasbru commented on pull request #13876: KAFKA-10733: Clean up producer exceptions

2023-06-30 Thread via GitHub
lucasbru commented on PR #13876: URL: https://github.com/apache/kafka/pull/13876#issuecomment-1614928567 @jolshan This makes sense. It seems to me that I need to go through the whole KIP and have an end-to-end solution before addressing these kinds of consistency problems. At least for the

[GitHub] [kafka] lucasbru commented on pull request #13876: KAFKA-10733: Clean up producer exceptions

2023-06-28 Thread via GitHub
lucasbru commented on PR #13876: URL: https://github.com/apache/kafka/pull/13876#issuecomment-1611471103 @jolshan Yes, the main reason for wrapping was consistency. However, I'm now considering a different kind of consistency - never wrap fatal errors - which is what was originally

[GitHub] [kafka] lucasbru commented on pull request #13876: KAFKA-10733: Clean up producer exceptions

2023-06-21 Thread via GitHub
lucasbru commented on PR #13876: URL: https://github.com/apache/kafka/pull/13876#issuecomment-1600924404 > 1. We might end up breaking customer's application code for exception handling with this change. What is our plan to prevent the inadvertent impact to customer's code on upgrade? >