atharvalade commented on PR #3103:
URL: https://github.com/apache/iggy/pull/3103#issuecomment-4524426822

   The biggest thing I caught was a credential leak where the derived Debug on 
S3Sink would dump the full AWS access key and secret key into logs whenever the 
struct got printed. I replaced that with a manual Debug impl that just shows 
the bucket name and added a regression test so it never sneaks back in. On the 
metrics side, the received count was being incremented after processing the 
batch, which meant if something failed halfway through you could lose track of 
how many messages actually got dropped. I moved that counter up front and added 
logic to correctly mark the unprocessed remainder as lost. I also cleaned up 
the dependency situation since `once_cell` was listed but never used, 
`simd-json `was only needed in tests but lived in regular deps, and the 
`cargo-machete` ignore list was papering over all of it. 
   
   The `max_retries` field got renamed to `max_attempts` with a serde alias so 
existing configs still work, and the README had a contradiction where it said 
timestamp was wall clock time when the code actually derives it from the first 
message in the buffer, so I fixed the docs to match reality. For the lost 
marker file that records data gaps, I routed it through the retry logic so a 
transient S3 hiccup does not silently eat your gap record. Then I wrote about 
twenty unit tests covering the pure functions like retry status classification, 
flush payload extraction, config validation, and the Debug redaction. Finally I 
stood up a full integration test suite using MinIO in a container that 
validates the sink actually writes `jsonl` to S3 with the correct path layout 
and rotates files properly when the message count threshold is hit.
   
   /ready


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to