I used only the Debian Etch package manager, nothing compiled myself.
Now, on a testapp when I do a script/server, I get the:
** Ruby version is not up-to-date; loading cgi_multipart_eof_fix
warning. But it’s not serving my app. Ruby is immediately on 100% CPU
load (and 0.7% memory) and stays there.
When I press CTRL-C on the console I started the server from, I get
=> Booting Mongrel (use ‘script/server webrick’ to force WEBrick)
=> Rails application starting on http://0.0.0.0:3000
=> Call with -d to detach
=> Ctrl-C to shutdown server
but still, nothing is served and Ruby process at 100% CPU.
Trying to do a mongrel_rails -B to debug it produces no debug log. The
hang is very immediate.
did you tried to use mongrel from backports.org (1)? Mongrel Debian
package
removes this fix because Debian’s ruby already has this included.
Just add backports.org entry to your /etc/apt/sources.list file:
deb http://www.backports.org/debian etch-backports main contrib non-free
I’ve found that mongrel tries to require some library repeatedly,
because it gets an unrelated exception while trying to load it. For me,
the missing library was actually fastthread, and installing it solved
the problem.
To find out what is the actual error, add a print statement after line 8
in lib/mongrel/gems.rb.
Hope that helps,
Paolo, you can confirm this behavior happens also with “mongrel_rails”
and not just script/server?
–
Luis L.
Multimedia systems
A common mistake that people make when trying to design
something completely foolproof is to underestimate
the ingenuity of complete fools.
Douglas Adams
I’ve found that mongrel tries to require some library repeatedly,
because it gets an unrelated exception while trying to load it. For me,
the missing library was actually fastthread, and installing it solved
the problem.
To find out what is the actual error, add a print statement after line 8
in lib/mongrel/gems.rb.
Hope that helps,
Paolo, you can confirm this behavior happens also with “mongrel_rails”
and not just script/server?
Yes, it did happen. After installing fastthread, I can’t reproduce it
anymore if I uninstall it. mongrel_rails now exits with a clean
backtrace when started without fastthread installed, and script/server
runs webrick instead.
I don’t know what exactly was causing the problem… maybe a wrong
installed version of fastthread? The exception caught in
lib/mongrel/gems.rb:8 was:
‘no such file to load – fastthread’ - sorry, I didn’t save the
backtrace.
I just had the same exact problem with the same solution. The wrong
platform
of fastthread was installed, which was causing mongrel to spin. That
should
probably dump with a stacktrace instead of spinning, I’ll see if I can
learn
enough mongrel to write a patch.
I used only the Debian Etch package manager, nothing compiled myself.
Now, on a testapp when I do a script/server, I get the:
** Ruby version is not up-to-date; loading cgi_multipart_eof_fix
warning. But it’s not serving my app. Ruby is immediately on 100% CPU
load (and 0.7% memory) and stays there.
When I press CTRL-C on the console I started the server from, I get
=> Booting Mongrel (use ‘script/server webrick’ to force WEBrick)
=> Rails application starting on http://0.0.0.0:3000
=> Call with -d to detach
=> Ctrl-C to shutdown server
but still, nothing is served and Ruby process at 100% CPU.
Trying to do a mongrel_rails -B to debug it produces no debug log. The
hang is very immediate.
I’ve found that mongrel tries to require some library repeatedly,
because it gets an unrelated exception while trying to load it. For me,
the missing library was actually fastthread, and installing it solved
the problem.
To find out what is the actual error, add a print statement after line 8
in lib/mongrel/gems.rb.
Hope that helps,
Paolo C.
This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.
