My situation is this:
Each time before uploading a war file to a web application folder, I stop Tomcat by calling sh shutdown.sh. It took about 30 seconds to shut down completely. But now it does not work well.
Actually, it worked, because when I access the application from a web page, it gives an error 503 (in the "Maintenance" section). But when I use ps aux | grep tomcatfor verification, the tomcat process still exists. And he will be there for about 5-10 minutes.
I understand that it may take extra time to complete all tasks, but it is too slow (5-10 minutes) before it completely stops. I do not understand why this is happening, but there must be some reason. Maybe something has to do with the code or with the new script deployment we used recently. I just barely know where to check.
This is important for our team, because we use "automatic deployment", in which we use a script to automatically fill the war file, load and deploy at a specific time. If we start a new tomcat instance before the old one is successfully completed, it will hang there forever, and the kill -9 cleanup task is complicated.
Is there anyone who has experimented with this problem? Any key would be appreciated.
source
share