Aias00 opened a new issue, #6546:
URL: https://github.com/apache/shenyu/issues/6546

   ## Description
   Within a selector group (grouped by `contextPath` or `appName` in 
`buildData`), the URIs are split only by `EventType`. For each resulting list, 
the code picks **one** namespace via `findFirst()`: 
`register.stream().map(URIRegisterDTO::getNamespaceId).filter(...isNotBlank).findFirst().ifPresent(namespaceId
 -> service.registerURI(selectorName, register, namespaceId))`. Every URI in 
`register` is then registered under that single namespaceId, ignoring each 
DTO's own `namespaceId`. The same pattern is used for `heartbeat` and `offline`.
   
   ## Location
   ```
   
shenyu-admin/.../disruptor/subscriber/URIRegisterExecutorSubscriber.java:88-105
   ```
   
   ## Impact
   Two namespaces share the same `contextPath`/`appName` and a published batch 
contains URIs from both namespaces -> cross-namespace URI mis-registration: 
URIs belonging to namespace B are recorded under namespace A, leading to 
routing/selectors being created in the wrong namespace. `offline` is especially 
dangerous — taking a backend offline in the wrong namespace while leaving it 
"online" in the correct one. `RegisterServerConsumerExecutor.isValidData` only 
checks that each URI's `namespaceId` is non-blank — it does not enforce 
uniformity.
   
   ## Suggested fix
   Sub-group by `namespaceId` before dispatching: 
`uriList.stream().collect(Collectors.groupingBy(URIRegisterDTO::getNamespaceId))`
 and call `service.registerURI/heartbeat/offline` once per `(namespaceId, 
subList)`.
   
   ## Related existing issue(s)
   None
   
   _Identified during the 2026-08-02 audit; full list in 
[`docs/issue-candidates-2026-08-02.md`](docs/issue-candidates-2026-08-02.md)._


-- 
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