What steps will reproduce the problem?
1. Do a clean deploy on a security heavy system (freeBSD)
2. Start the application under Passenger
3. Check the output, passenger spawns crash due no log file.
What version of RubyCAS-Server are you using? How is it installed (rubygem,
manual install)? How are you running it (webrick, mongrel, passenger,
etc.)?
We use the source from git with a few minor tweaks to make the application
runnable under Passenger. This means a few tweaks to the structure of the
project to suit our needs and prodiving a config.ru file to set up the
environment. The code itself is not touched.
If relevant, please paste your RubyCAS-Server config.yml file here.
The config file is not relevant. Just make sure it is a valid config file, and
that you output everything to specific log file in for example the log
directory inside the project.
Please provide any additional information below.
When you start up the application and the log files do not exist, your
Passenger instances will simply crash when booting up the application. The
error messages you receive say that there is no such file => pointing to
whatever you set your log file.
This does not happen on every configuration. So far I have seen it happen on
all our FreeBSD boxes during deployment, our TeamCity agents, which run Debian.
There the feature tests of cucumber fail with this error message.
On our development systems (Mac OS Lion) this does not happen at all.
I'm guessing this has something to do with permissions under various Linux
systems. I've tried fiddling with the code regarding to the Logger, but
whatever I tried, nothing resolved this problem.
You can bypass this on deployment by running the "touch" command as the same
user that will be running the application and make sure the file exists.
For the TeamCity tests, I have not found a way around this yet.
Original issue reported on code.google.com by arne.de....@gmail.com on 13 Aug 2012 at 4:53
Original issue reported on code.google.com by
arne.de....@gmail.comon 13 Aug 2012 at 4:53