Hi list,
So I’ve been trying to upgrade from 0.6.7-svn to the 0.6.9 gem using a
Site5 account, and something’s blocking me. I am running gems 1.1.0,
ruby 1.8.7 and have the radiant 0.6.9 gem installed locally with the
binary in my $PATH. Everything runs out of my home directory, including
the ruby executable.
I was a little unsure of how to switch from svn to gem, so I installed
gem radiant to a new directory using simply “radiant .” from within that
dir. I then copied over the .htaccess, database.yml, and vendor/, and
merged the custom lines of the environment.rb file. I checked
permissions on the public directory, and everything was identical.
I then ran ‘rake production db:migrate’ and ‘rake radiant:update’ and
updated all the extensions. I tested using ‘ruby public/dispatch.fcgi’
in bash, and it spit back the html I was looking for without a hitch.
Finally, I deleted my public_html symlink to the old svnradiant, and
made a new public_html symlink to the new gemradiant. I pkilled the
distpatch.fcgi’s, and then… nothing: 30 second delay followed by a
500 error. My error log in siteadmin just said “process failed to stay
running after n attempts, etc.”, but nothing more specific than that.
After some frustration, I switched the public_html symlink back to my
0.6.7-svn public directory, and dispatch.fcgi caught again, no
problems. This is even using the post-migration database.
Has anyone had an experience like this? Is there some crucial step I
could be missing here? Some thing just won’t work. My error logs are
little help.
Here’s the recurring production.log entry:
Processing SiteController#show_page (for at 2008-08-05 16:17:35) [GET]
Parameters: {“url”=>"/", “action”=>“show_page”, “controller”=>“site”}
Completed in 1.20128 (0 reqs/sec) | DB: 0.00000 (0%) | 200 OK [http://:]
And here’s my fastcgi.crash.log:
[05/Aug/2008:14:27:22 :: 15693] starting
[05/Aug/2008:14:27:23 :: 15693] stopping gracefully
[05/Aug/2008:14:29:07 :: 16043] starting
[05/Aug/2008:14:29:07 :: 16043] stopping gracefully
[05/Aug/2008:16:17:35 :: 9096] starting
[05/Aug/2008:16:17:36 :: 9096] stopping gracefully
[…etc…]
I’m not sure what I could be missing here, but I’m hoping someone has
had a similar experience, or knows another way to help me diagnose this
stealth error.
Thanks everybody,
–danny