SEPURI-SAI-KRISHNA commented on PR #13261:
URL: https://github.com/apache/seatunnel/pull/13261#issuecomment-6909923438

   @SEZ9 Thanks. Your notes refer to `24b8e5153a`; the current head is 
`50e1c1087a`.
   
   1. **F1**: since `24b8e5153a`, `run()`, `addSplitsBack`, `registerReader` 
and `assignSplit` are all `synchronized`, so the check and the assignment 
cannot interleave. The concurrency test goes on only once the other thread is 
blocked on the enumerator; 10 of 10 runs pass, and without `synchronized` 10 of 
10 fail.
   2. **F2**: the Zeta order is 
`testRestoredSplitIsAssignedOnceWhenTheReaderRegistersAndRunFollows` (one 
assignment, no second `new SingleSplit(null)`), and `run()` after a 
re-assignment is in 
`testSplitGivenBackWhileTheReaderIsRegisteredIsAssignedAtOnce`. On a real 
engine, `HttpIT` restarts single split jobs on Flink, including one that 
completes with its exact row after two restarts; the description now says how 
this was checked.
   3. **F4**: the Javadoc says why an empty list can be ignored: the reader 
restores its split from its own checkpoint state.
   4. **F7**: the test class and methods are `public`, and `50e1c1087a` adds a 
debug log on the take-back path, as in the sibling enumerators.
   


-- 
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