Reduce size of buffer used to assemble batches #1204
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Motivation
The current implementation is using 128KB buffers to accumulate batches before sending. (128Kb is the max batch size, so the intention here is to avoid reallocations on that buffer as it grows).
The problem starts with (batching + “many” partitions + slow responses from broker, due to network limit or reloads). At that point, the batches are still being prepared “on-time”, eg: every 1ms. With many partitions, the per-partition rate is comparatively low, so we end cutting a batch per each message and queuing them for sending. But each message now is taking 128Kb direct memory, instead of 10K or 1K..
Modifications
Adjust the initial buffer size to the max batch size previously sent out by this particular producer.