I have an import-script, that is going for about 45mins. In lucee 5 i always used requesttimeout=“0”, so that the script doesn’t time out. Now in lucee 6 it times out after 12mins although I used requesttimeout=“0”. Is this an intentional change, because i didn’t find anything about it.
What are the timeout settings in NGINX and Tomcat?
If not set, these have defaults.
Though I am pretty sure that the defaults are not 12 minutes…
Are you sure the operation is stopping?
Or has it just reported the timeout to the browser?
We have a process that takes over an hour.
The browser returns after 10 minutes with a timeout or bad gateway or similar error - because NGINX has a timeout of 10 minutes.
BUT our process is still working in the background.
If you don’t already - add some logging to your task - and review the log AFTER the timeout - to see if it is still running or not.
Also it is helpful if you let us know what version exactly of Lucee you’re using…
Did you use Commandbox to start it / which version?
Does it use Tomcat - which version?
What version of Java is used by Lucee (not necessarily the same as what the underlying server has / uses.
Nginx has keepalive_time setting you can configure, but i have a feeling this is a tomcat issue
I most certainly do not pretend to be “right”…
As Tomcat controls Lucee - somewhat - wouldn’t a tomcat timeout stop the lucee process from running, too?
Or is that (Tomcat timeout) - a definite - separate layer above the Lucee Servlet container?
Because the Lucee processing - continues - despite the timeout, displayed in the bvrowser?
Gavin,
My understanding is Lucee explicitly extends tomcat. Tomcat is the ultimate authority in the relationship.
As for Nginix, it defaults to an hour timeout. So now back to looking at the application configuration and tomcat settings. Maybe another weird error with localhost setting using the string “localhost” versus the ip address of 127.0.0.1 or ::1
1 Like