[
https://issues.apache.org/jira/browse/KAFKA-5032?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16026211#comment-16026211
]
Ismael Juma commented on KAFKA-5032:
------------------------------------
[~apurva], docs sound good. Maybe we should consider changing the config
default as well. If that makes sense, it's a one liner.
> Think through implications of max.message.size affecting record batches in
> message format V2
> --------------------------------------------------------------------------------------------
>
> Key: KAFKA-5032
> URL: https://issues.apache.org/jira/browse/KAFKA-5032
> Project: Kafka
> Issue Type: Sub-task
> Components: clients, core, producer
> Reporter: Ismael Juma
> Priority: Critical
> Labels: exactly-once
> Fix For: 0.11.0.0
>
>
> It's worth noting that the new behaviour for uncompressed messages is the
> same as the existing behaviour for compressed messages.
> A few things to think about:
> 1. Do the producer settings max.request.size and batch.size still make sense
> and do we need to update the documentation? My conclusion is that things are
> still fine, but we may need to revise the docs.
> 2. Consider changing default max message set size to include record batch
> overhead. This is currently defined as:
> {code}
> val MessageMaxBytes = 1000000 + MessageSet.LogOverhead
> {code}
> We should consider changing it to (I haven't thought it through though):
> {code}
> val MessageMaxBytes = 1000000 + DefaultRecordBatch.RECORD_BATCH_OVERHEAD
> {code}
> 3. When a record batch is too large, we throw RecordTooLargeException, which
> is confusing because there's also a RecordBatchTooLargeException. We should
> consider renaming these exceptions to make the behaviour clearer.
> 4. We should consider deprecating max.message.bytes (server config) and
> message.max.bytes (topic config) in favour of configs that make it clear that
> we are talking about record batches instead of individual messages.
> Part of the work in this JIRA is working out what should be done for 0.11.0.0
> and what can be done later.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
