[PR] [fix] ensure that when acceptor is stopped it doesn't log an error for nothing [tomcat]

2025-05-17 Thread via GitHub
rmannibucau opened a new pull request, #857: URL: https://github.com/apache/tomcat/pull/857 Proposal to have a clean shutdown (follow up of the mail on the list). This only handles the case bindOnInit=false. -- This is an automated message from the Apache Git Service. To respond to

Re: Clean stop of Tomcat (embed)

2025-05-17 Thread Romain Manni-Bucau
ps: it is in bindoninit=false, if true there seems to be a mismatch between stop and destroy but the log is a similar warning (The acceptor thread [http-nio-xxx-Acceptor] did not stop cleanly) Romain Manni-Bucau @rmannibucau | .NET Blog

Clean stop of Tomcat (embed)

2025-05-17 Thread Romain Manni-Bucau
Hi all, When stopping tomcat (stop()), it does call org.apache.catalina.core.StandardService#stopInternal which tries a graceful shutdown on the protocol->endpoint so we end up in org.apache.tomcat.util.net.NioEndpoint#doCloseServerSocket but the endpoint is in running=true state so acceptor threa