SEZ9 commented on PR #11512:
URL: https://github.com/apache/seatunnel/pull/11512#issuecomment-7023259774

   @goutamadwant thanks for the point-by-point mapping. You didn't ask a 
question, so here is where each of F1–F8 stands for me. I have not re-checked 
any of this against the code, so none of them are closed on head `738c0acc46` 
yet.
   
   - **F1 / F8:** Removing the per-split detail list, so the enumerator report 
carries counts only and there is no `CdcSnapshotSplitProgress` type, resolves 
both by removal. They stay open until the rework is pushed on #13239, and until 
this PR and #13240 drop the serializer block and the assigner detail code.
   - **F2:** The "Positions" wording (offset coordinates only, never 
credentials, connection URLs or authentication material) is what I was asking 
for. Please make sure it survives in the reworked doc.
   - **F3:** Writing owner, lifecycle, assignment status and accuracy by 
`name()` and rejecting unknown names in `CdcProgressReportSerializer` is the 
right fix. I'll confirm it on the re-review, including the phases-as-strings 
addition.
   - **F4:** This stays open until the "Runtime collection in Zeta" section in 
`docs/*/developer/cdc-progress.md` is updated for the new names. It should 
still cover both the worker pull for readers and the coordinator-owned 
enumerator path.
   - **F5:** Limiting this to `connector-cdc-base` and the connectors built on 
it is fine. It stays open until the known-gap line is in the API doc.
   - **F6:** Agreed that it is not fixed yet. The `SNAPSHOT` Javadoc needs the 
wording you describe: the consumed position is `UNAVAILABLE` during `SNAPSHOT`, 
and the reader does not infer enumerator state.
   - **F7:** One thing I need from you here. You say the constructors document 
the non-negative and `assigned == completed + running` (only when all `EXACT`) 
rules. Are those rules enforced in the constructor, or only described in 
Javadoc? The finding was about missing validation, so please confirm how the 
single reworked constructor (with `inFlight`) handles invalid values.
   
   Remaining asks: push the rework, answer the F7 question, and add the F5 and 
F6 doc changes. Then ping me and I'll re-review F1–F8 against the revised 
commits.
   
   <!-- streview-comment:2458 -->


-- 
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