I've just upgraded Cherokee to 1.2.100 from 1.2.99 and have noticed that my previously-configured Redmine installation (originally from the Cherokee Market) does not spawn when accessed. Specifically, under 1.2.99 the local Ruby interpreter would be spawned upon someone accessing a given directory, and using the same configuration on the same host (RHEL 6, 64 bit), the Ruby instance does not get spawned. The source is configured thusly:
source!1!env!GEM_HOME = /opt/redmine/.gem source!1!env!GEM_PATH = /opt/redmine/.gem:/usr/lib/ruby/gems:$GEM_PATH source!1!env!RAILS_ENV = production source!1!env!RUBYLIB = /opt/redmine/lib:$RUBYLIB source!1!env_inherited = 0 source!1!host = 127.0.0.1:56330 source!1!interpreter = /usr/bin/ruby /opt/redmine/redmine/script/server thin start --binding=127.0.0.1 --port=56330 --environment=production source!1!nick = Redmine source!1!timeout = 3 But accessing the given URL sees nothing logged in the Cherokee error logs. I've also tried adjusting the timeout to 30 seconds with no difference; as best I can tell, no Ruby process gets started at all. That said, running the command manually works without a problem: sudo -u cherokee RUBYLIB=/opt/redmine/lib:$RUBYLIB GEM_HOME=/opt/redmine/.gem RAILS_ENV=production GEM_PATH=/opt/redmine/.gem:/usr/lib/ruby/gems:$GEM_PATH /usr/bin/ruby /opt/redmine/redmine/script/server thin start --binding=127.0.0.1 --port=56330 --environment=production so there doesn't appear to be any trouble with the interpreter actually starting (and afaict, the interpreter doesn't daemonize or detatch itself). For now, I've taken to starting the instance myself manually in a screen session, and Cherokee is able to use this without problem. However, this means manually intervening. Does anyone have any thoughts? Thanks in advance! -- David
_______________________________________________ Cherokee mailing list [email protected] http://lists.octality.com/listinfo/cherokee
