Aias00 commented on PR #4385:
URL: 
https://github.com/apache/rocketmq-dashboard/pull/4385#issuecomment-5756821342

   Withdrawing this PR based on the review findings.
   
   RocketMQ Proxy 5.5.0 does not expose an independent lag source here:
   `ConsumerManagerActivity` forwards `GET_MAX_OFFSET` and 
`QUERY_CONSUMER_OFFSET` to the same
   Broker. Since the fallback starts only after `examineConsumeStats` has 
already read Broker offsets,
   the extra calls cannot justify their latency or RPC volume without a real 
deployment capture.
   
   I prototyped the requested 8-way bounded concurrency, shared three-second 
deadline, queue
   deduplication, and partial `UNKNOWN` degradation locally. The focused suite 
passed 129 tests, but
   that only proves the fan-out can be bounded; it does not prove the fan-out 
is useful. The synthetic
   `broker=-1 / proxy=50` fixture is not acceptable benefit evidence because 
both paths ultimately
   read the same Broker.
   
   Issue #4367 has been revised to document the upstream constraint and 
reopening criteria. #4386 keeps
   ownership of unavailable-lag representation, and #4387 remains independent 
Ops/runtime work.


-- 
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