zjncs opened a new pull request, #11132:
URL: https://github.com/apache/rocketmq/pull/11132
### Motivation
The `-c` option of `mqadmin checkRocksdbCqWriteProgress` was only used for
an existence check, while the request loop afterwards walked the whole
`brokerAddrTable`, so the check request was fired at brokers of every cluster —
not just the requested one. Brokers whose master was down resolved to a null
address, which `defaultMQAdminExtImpl` silently routed through the name server
channel.
### Modifications
Iterate the brokers of the requested cluster only, and skip brokers without
a master.
### Verification
Fail-before (new test, run against the unpatched code):
```
CheckRocksdbCqWriteProgressCommandTest#testCheckOnlySpecifiedCluster
java.lang.RuntimeException: CheckRocksdbCqWriteProgressCommand command failed
at
CheckRocksdbCqWriteProgressCommand.execute(CheckRocksdbCqWriteProgressCommand.java:99)
(check request hit a master-less broker outside the requested cluster)
```
Pass-after:
```
Tests run: 2, Failures: 0, Errors: 0, Skipped: 0 --
CheckRocksdbCqWriteProgressCommandTest
```
--
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]
