2

We have set the default error of our host to 503, therefore our external systems get a "503 Service Unavailable" during deployment. This is interpreted just like as the service was not running and the request is retried.

This works as intended.

The shutdown behaves strangely. No matter if we stop Wildfly10 with SIGTERM or via the CLI shutdown command, we first see the expected HTTP503s for 1-2 Seconds but then we see HTTP404 for a few seconds until the server is completely shut down.

I think the Undertow subsystem should not allow any new connections, but if it does I expect it to respond with our default error (503)...

What are we doing wrong?

Martin Rauscher
  • 1,700
  • 1
  • 14
  • 20
  • how do you call :shutdown command in cli? – Tomaz Cerar Mar 03 '16 at 20:22
  • Better way to do this is to have smart load balancer/reverse proxy, have the balancer redirect to the active server which is up/ it should do that automatically. – Zeus Apr 11 '17 at 20:30

0 Answers0