csongs

就我自己看代碼,設計上來說的確如您所說的 不能把一個app分給兩個server來分擔

他的設計就是可以多個server同時進行, 簡單說當多個server時,會選舉出負責處理的server leader 另外不同app(執行應用)不一定要在同一個server 作者也有進行說明: [連結](https://mp.weixin.qq.com/s?__biz=MzA5MzYyNzQ0MQ==&mid=2247488202&idx=1&sn=129f0e65ee1d5104e2db391e4ef67e50&chksm=905ba794a72c2e824b9ebc642eb55af1ec6215144f0f660c2a6163c3aafe10d606835edacbe3&scene=178&cur_album_id=1424305043590791168#rd)

I use hazelcast=5.2.2 and vertx=4.4.0 is work But Use hazelcast=5.3.5 and vertx=4.5.7 is not work I deploy on K8S 1.26 Use KUBERNETES_API This is log ``` 2024-04-22 10:00:31 INFO [vert.x-worker-thread-0]...

I find it... - k8s deployment - before ```yaml ports: - containerPort: 8080 - containerPort: 5701 ``` - after ```yaml ports: - containerPort: 5701 - containerPort: 8080 ``` To run...

動態註冊worker的話 ,我自己是用 hazelcast去實現, 不過我是架在k8s (hazelcast 已有內建跟k8s做整合) 目前這方案就是無法跨k8s cluster