It is exactly what I mean. And I am sure many users would get benefits from the event system, especially the users want a observable system based on APISIX. I think it worth to be invested at some time. And the designs you considered are also what I think about recently. It would be a pleasure that the Community will be considering my suggestion.
At 2022-11-27 14:30:59, "Chao Zhang" <zchao1...@gmail.com> wrote: >Hi, > >Thanks for your suggestion. It sounds like building an event system. So >roughly we need do the things below: > >1. Design all events we need >2. Design how we collect them and maintian them temporarily >3. Design how we expose them >4. Design how we visualize and analyze them. > >赵旭 <yunbaochen...@163.com>于2022年11月26日 周六17:02写道： > >> Hello, Community, as we known the state of resources(or some important >> variable) changes in runtime.Such as the nodes in upstream may >> >> become from healthy to unhealthy, the router radix will rebuild when the >> routes change. I think it is a good idea >> >> to offer an ability that the APISIX can report some important change >> events to DashBorad.When users have doubts about the >> >> apisix runtime state, the reports on DashBorad(or some other place) will >> be good proofs. >> >> >> Best regards. >> >> >> xu.zhao > >-- >Best regards >Chao Zhang > >https://github.com/tokers
