Idempotent writes with writer ID allocation and batch sequencing #389
Description
Search before asking
- I searched in the issues and found nothing similar.
Description
The client retries failed batches on network errors up to i32::MAX times by default. Without idempotence, the server has no way to detect retried batches, when writer_id is -1 (current default), the server skips all dedup checks and appends unconditionally. A timeout where the server received the batch but the ack was lost results in silent duplicate records.
Willingness to contribute
- I'm willing to submit a PR!