Hi Andrew Schofield and All: Sorry to trouble you ! It seems that the last mail is lost in the community email group. I paste it here and resend it again. BTW: I already started the vote here: https://lists.apache.org/thread/mkdm421nrdwc9z2qtsyom3xz25skdfmr If you think it is ok without any concern for you. Can you help to vote? Thanks a lot.
At the same time, any suggestions are welcome, and a vote does not mean the discussion is closed. The KIP Link: https://cwiki.apache.org/confluence/display/KAFKA/KIP-1235%3A+Correct+the+default+min.insync.replicas+to+2+for+the+__remote_log_metadata+topic the email content lost in the thread: *'''* *Hi Andrew:Thank you very much for joining the discussion and providing your comments. I feel this KIP shouldn’t be very controversial since it follows a standard practice and kafka already have internal topic practices to reference. So I’m thinking of leaving the discussion open for another week, and then we can probably start/continue the voting. This also exceeds the one-month discussion period to ensure everyone has had a chance to see it.Thanks again. RegardsJian* *''''* jian fu <[email protected]> 于2025年12月3日周三 17:19写道： > Hi Andrew: > > Thank you very much for joining the discussion and providing your > comments. I feel this KIP shouldn’t be very controversial since it follows > a standard practice and kafka already have internal topic practices to > reference. So I’m thinking of leaving the discussion open for another week, > and then we can probably start/continue the voting. This also exceeds the > one-month discussion period to ensure everyone has had a chance to see it. > > Thanks again. > > Regards > Jian > > Andrew Schofield <[email protected]> 于2025年12月3日周三 01:46写道： > >> Hi Jian, >> Thanks for the KIP. This looks like a sensible idea to improve the >> default configuration for this topic. >> >> Thanks, >> Andrew >> >> > On 14 Nov 2025, at 05:27, jian fu <[email protected]> wrote: >> > >> > Just bumping this thread for more discussion. >> > >> > Thanks! >> > >> > jian fu <[email protected]> 于2025年11月5日周三 14:54写道： >> > >> >> Hi everyone, >> >> >> >> I'd like to start a discussion on KIP-1235, a small change to set >> >> min.isr=2 for the __remote_log_metadata topic. >> >> >> >> KIP: >> >> >> https://cwiki.apache.org/confluence/display/KAFKA/KIP-1235%3A+Correct+the+default+min.insync.replicas+to+2+for+the+__remote_log_metadata+topic >> >> >> >> Problem: >> >> >> >> Currently, __remote_log_metadata relies on the broker default for >> min.isr >> >> (typically 1). >> >> With min.isr=1 and factor=3, acknowledged writes can be lost if the >> leader >> >> crashes before replication >> >> completes, making remote data permanently unreachable. >> >> >> >> What's more, the similar topic: __transaction_state sets min.isr=2 via >> >> transaction.state.log.min.isr. >> >> We should provide the same durability for __remote_log_metadata. >> >> >> >> >> >> Proposed Change: >> >> >> >> Add one new broker config: remote.log.metadata.topic.min.isr (default: >> 2) >> >> >> >> When creating __remote_log_metadata, TopicBasedRemoteLogMetadataManager >> >> will set >> >> min.insync.replicas=2 at the topic level. The change only affects new >> >> clusters enabling Tiered Storage. Existing topics are unchanged. >> >> >> >> >> >> Looking forward to your feedback! >> >> >> >> Best regards, >> >> Jian >> >> >> >> > > > >
