Corner Rat having pathname trouble in boot.rb

mperham · May 19, 2006, 3:43pm

Hi folks,

I was on the typo mailing list about a year ago, but I seem to be having
problems again. I’m not sure if it’s just that Dreamhost is crappy for
RoR,
but on the whole, I’ve found running Typo a very frustrating experience.
(you know, apart from the occasional periods where it would run
perfectly
fine and be terrific… I really appreciate all the hard work the devs
have
put into making Typo great, I just wish I could get it going more
reliably
for myself.)

Anyhow, I upgraded to r1053, which seemed to be okay, except for a bug
where
saving an article automatically published it. I tried ‘svn up’-ing to
r1054,
which broke everything, so I downgraded again. Everything had seemed
fine
last night, but now I’m getting 500s in the browser, and this on the
command-line:

[foothill]$ ./dispatch.fcgi
./…/config/boot.rb:7:in require': no such file to load -- pathname (LoadError) from ./../config/boot.rb:7 from ./../config/environment.rb:8:inrequire’
from ./…/config/environment.rb:8
from ./dispatch.fcgi:21:in `require’
from ./dispatch.fcgi:21

Any input or suggestions would be welcome.

Mike

mperham · May 19, 2006, 4:33pm

On Fri, 2006-05-19 at 09:40 -0400, Mike P. wrote:

I was on the typo mailing list about a year ago, but I seem to be
having problems again. I’m not sure if it’s just that Dreamhost is
crappy for RoR, but on the whole, I’ve found running Typo a very
frustrating experience. (you know, apart from the occasional periods
where it would run perfectly fine and be terrific… I really
appreciate all the hard work the devs have put into making Typo great,
I just wish I could get it going more reliably for myself.)

Just checking: rails is frozen in vendor/ right? If not, then that’s
why it’s so unstable. Other than the latency from their machines being
so drastically overcommitted, personally I don’t find DH to be too bad.

I feel your pain. I’m wondering if Typo is starting to pull a
Debian…? The stable version is hopelessly out of date and the
development version is often breaking in new and strange ways, leaving
no version suitable for production usage. Just a thought.

- Scott

mperham · May 19, 2006, 4:49pm

Hi Scott,

Thanks for your reply. How would I determine if Rails is frozen?

I observe that when I run the svn upgrade, it pulls in a specific
revision
of Rails and places it in that directory, but how do I check if that’s
what’s actually being used?

I know that at various times, I’ve installed my own ruby and rails,
although I was fairly sure I’d been running Dreamhost’s ruby and
Dreamhost’s
rails. Indeed a “whereis rails” from my shell shows “/usr/bin/rails”. On
the
other hand, when I navigate to /usr/bin, there’s no rails to be found…

Is it just a matter of (somewhere) pointing that missing pathname
variable
to vendor/rails?

Mike

mperham · May 19, 2006, 5:16pm

p.s. I was running an old version (frozen rails 1.0)… so I backed up
and updated to HEAD of trunk and am still getting the same problems…
with frozen rails 1.1.

mperham · May 19, 2006, 5:10pm

I’m having trouble as of today on DH. Blog was running fine yesterday…

When I try to run script/console I get

[scagnetti]$ ./script/console
./script/…/config/boot.rb:7:in require': no such file to load -- pathname (LoadError) from ./script/../config/boot.rb:7 from ./script/console:2:inrequire’
from ./script/console:2

And in my access.log I get this.

[Fri May 19 08:04:20 2006] [error] [client 71.39.198.114] FastCGI: comm
with (dynamic) server
“/home/thoughtstoblog/sites/thoughtstoblog.com/public/dispatch.fcgi”
aborted: (first read) idle timeout (120 sec)
[Fri May 19 08:04:20 2006] [error] [client 71.39.198.114] FastCGI:
incomplete headers (0 bytes) received from server
“/home/thoughtstoblog/sites/thoughtstoblog.com/public/dispatch.fcgi”

mperham · May 19, 2006, 5:28pm

It looks like a problem with Ruby on a few servers at DH. I emailed
support using their OMG PEOPLE ARE GOING TO DIE severity…

Even running a simple gem list --local on the server results in this

[scagnetti]$ gem list --local
/usr/local/lib/site_ruby/1.8/rubygems.rb:1:in require': no such file to load -- rbconfig (LoadError) from /usr/local/lib/site_ruby/1.8/rubygems.rb:1 from /usr/bin/gem:3:inrequire’
from /usr/bin/gem:3

mperham · May 19, 2006, 5:19pm

Cool, glad to know it’s not all me. Once I couldn’t even “rake migrate”
from
the prompt, I wondered if maybe DH had pulled another surprise upgrade
on
us.

I submitted a support request about an hour ago; haven’t heard anything.

Mike

mperham · May 19, 2006, 5:35pm

On May 19, 2006, at 7:48 AM, Mike P. wrote:

Thanks for your reply. How would I determine if Rails is frozen?

I observe that when I run the svn upgrade, it pulls in a specific
revision of Rails and places it in that directory, but how do I
check if that’s what’s actually being used?

Use the script/about command in your Rails project. It will tell you
the versions of Rails and all the gems being used. If you’re running
off trunk it shows the svn version number as “Edge Rails revision”.

–
Josh S.
http://blog.hasmanythrough.com

mperham · May 19, 2006, 5:41pm

Ah, well then, it’s definitely a rails install gone AWOL:

[foothill]$ ./script/about
./script/…/config/boot.rb:7:in require': no such file to load -- pathname (LoadError) from ./script/../config/boot.rb:7 from ./script/about:2:inrequire’
from ./script/about:2

How do I get it to use the Rails that’s inside /vender?

Mike

mperham · May 19, 2006, 5:38pm

On Fri, 2006-05-19 at 10:48 -0400, Mike P. wrote:

Hi Scott,

Thanks for your reply. How would I determine if Rails is frozen?

$ ls vendor/rails
actionmailer actionwebservice activesupport
actionpack activerecord railties

Any other result means that you will suffer every time DreamHost
upgrades its version of rails.

I observe that when I run the svn upgrade, it pulls in a specific
revision of Rails and places it in that directory, but how do I check
if that’s what’s actually being used?

Looks like you are frozen. Good. I don’t know how to verify that those
are indeed the versions being used but I wouldn’t worry about it much.

I know that at various times, I’ve installed my own ruby and rails,
although I was fairly sure I’d been running Dreamhost’s ruby and
Dreamhost’s rails. Indeed a “whereis rails” from my shell shows
“/usr/bin/rails”. On the other hand, when I navigate to /usr/bin,
there’s no rails to be found…

Is it just a matter of (somewhere) pointing that missing pathname
variable to vendor/rails?

Actually, my blog is down too. I’m getting errors about missing
standard Ruby modules in my logs and, as you noticed, the rails command
is MIA. Looks like DreamHost has broken Rails worldwide.

[holt]$ /usr/lib/ruby/ruby/gems/1.8/bin/rails --version
/usr/local/lib/site_ruby/1.8/rubygems.rb:1:in require': no such file to load -- rbconfig (LoadError) from /usr/local/lib/site_ruby/1.8/rubygems.rb:1 from /usr/lib/ruby/ruby/gems/1.8/bin/rails:9:inrequire’
from /usr/lib/ruby/ruby/gems/1.8/bin/rails:9

Not much you or I can do except investigate alternative hosting. Bogus.

- Scott

mperham · May 20, 2006, 12:25am

Even with Dreamhost having Rails back up, I still seem to be having some
difficulties. I’ve got a shell open tailing the dev log, and some
requests
are working, but others throw up stuff like this:

SystemExit in Articles#index

Showing themes/dime/views/articles/_article.rhtml where line #0
raised:

exit

Extracted source (around line #0):

1:

<%= article_link article.title, article %>

2:

Posted by <%=
author_link(article) %>
3: <%= js_distance_of_time_in_words_to_now article.created_at %>

Trace of template inclusion: /themes/dime/views/articles/index.rhtml

RAILS_ROOT: …/config/…
Application Trace http://uwmike.com/articles/page/3# | Framework
Tracehttp://uwmike.com/articles/page/3#| Full
Trace http://uwmike.com/articles/page/3#

#{RAILS_ROOT}/vendor/rails/railties/lib/fcgi_handler.rb:126:in exit' #{RAILS_ROOT}/vendor/rails/railties/lib/fcgi_handler.rb:126:in exit_now_handler’
/usr/local/lib/site_ruby/1.8/rubygems/custom_require.rb:18:in to_proc' #{RAILS_ROOT}/vendor/rails/railties/lib/fcgi_handler.rb:114:in call’

Request

Parameters: {“page”=>“3”}

Show session dump http://uwmike.com/articles/page/3#

Response Headers: {“cookie”=>[], “Cache-Control”=>“no-cache”}

mperham · May 20, 2006, 5:48pm

On Fri, 2006-05-19 at 08:30 -0700, Josh S. wrote:

On May 19, 2006, at 7:48 AM, Mike P. wrote:

Thanks for your reply. How would I determine if Rails is frozen?
Use the script/about command in your Rails project. It will tell you
the versions of Rails and all the gems being used. If you’re running
off trunk it shows the svn version number as “Edge Rails revision”.

But it doesn’t tell you if you’re running frozen gems from /vendor or
the globally-installed Rails.

Is there any way to determine this?

- Scott

mperham · May 19, 2006, 7:54pm

If you want a good revision of typo trunk, I’ve been having luck with
1039 and Rails 1.1.1.

Posting works, admin works, after some initial flakiness it has been
running for weeks.

mathew