All geek questions in one place

Fabric application application corruption

I just test the service fabric in place. I tested a lot of crash scenarios, but one of them I can not verify. How does SF work when node works well but the application crashes? For example, I have an unencumbered web api, and after one request it fails and shuts down (almost impossible, but this is only an assumption). SF needs to know about this, and in the next request to the same node, it must redirect the request to the same type of application hosted in any other node until the application starts again? I'm right? In statefull, it should do the same, but instead of redirecting to another node, should it use a replica?

I am trying to simulate this example using Restart-ServiceFabricDeployedCodePackage , but it probably reloads too fast and I cannot confirm my assumptions - I get a timeout.

+4
source share
1 answer

It is assumed that Service Fabric will redirect service requests, which is generally not true unless you use the built-in reverse proxy (you will find out if you use it because you will need to structure the request URLs in a certain way).

, -, IP: port. . .

SF . . SF , , , . node. node, . , .

+2

Source: https://habr.com/ru/post/1671115/

More articles:


All Articles