Thanks Bruno. I already have the following code in my environments/
production.rb:
# Use a different cache store in production
config.cache_store = :file_store, "#{RAILS_ROOT}/tmp/cache"
I'm still looking into this problem to see what I can figure out.
I've been going through my dev and production logs. Is there any way
to enable additional logging for a production server? I would really
like to track cache hits/misses/expirations on my production server,
but without the full stack trace shown in the development logs.
I'm creating another git branch to track this in so that I can
contribute any fixes upstream. I don't have any experience with unit
testing and may need some guidance in this area before my changes
could be pulled upstream.
Justin
On Mar 4, 5:47 pm, Bruno Bornsztein <[email protected]>
wrote:
> I'll have to take a closer look at this when I get more time, but could you
> try putting this in your environment.rb:
>
> ActionController::Base.cache_store = :file_store, "#{RAILS_ROOT}/tmp/cache"
>
> See if that solves your problems.
> More broadly, though, the caching and sweeping in CE needs a little TLC, and
> testing!
>
> Thanks,
> Bruno
>
> On Mon, Mar 2, 2009 at 10:32 PM, Justin G <[email protected]> wrote:
>
> > About a month ago I launched my community engine site into the wild.
> > At the time, I had some caching issues with the front page. I read up
> > on how caching works in rails and looked at the code a bit, but
> > everything looked OK. So I figured it was best to disable caching for
> > the site_index action and swore to revisit the issue at a later time.
>
> > I'm running my production site on mod_passenger so I also limited the
> > number of processes to 1. I figured that if the cache was being
> > stored in memory, then different processes could be serving from
> > different caches. Since the problem was hit or miss it seemed like
> > this could also be the issue. (And with a small site running on a VPS
> > with limited memory I wasn't yet worried about scalability.)
>
> > Anyway, I recently noticed a problem with comments that weren't
> > updating. I received an email that someone commented on my post, but
> > I could only see the comment if I was logged in. I figured it was a
> > caching problem so I restarted the server and the problem went away.
> > Here are some things I found while investigating the issue, and an
> > issue I am still having.
>
> > Original Problem
>
> > I tracked part of the problem down to the fact that the comment
> > sweeper wasn't clearing all copies in the cache. My development.log
> > file showed that the comment_sweeper was clearing /jtgeibel/posts/7-
> > testing (among other things), but it wasn't clearing /posts/show/7-
> > testing (the cache fragment that is actually being hit). I looked at
> > the code and made the following change:
>
> > - expire_action :controller => 'posts', :action => 'show', :id
> > => record.commentable , :user_id => record.commentable.user
> > + expire_action :controller => 'posts', :action => 'show', :id =>
> > record.commentable
>
> > From what I can tell, rails is only clearing the cache for the more
> > specific route containing a user_id. However, its the action posts/
> > show that is actually being cached. With this fix, I now see the
> > fragment cache being cleared in the log, but I'm still having one very
> > strange problem.
>
> > New Problem
>
> > I can see in the development log that the cache is being cleared, but
> > the first time I view the page (after logging out) I actually get a
> > cache hit. If I refresh the page again, I see the updated page. Here
> > is a commented portion of what I am seeing in my log file.
>
> > $ cat ../../../log/development.log |grep fragment
>
> > # ensure cache is full before posting a comment
> > Cached fragment hit: views/dev.pittsburghgardenexperiment.org/posts/
> > show/7-testing<http://dev.pittsburghgardenexperiment.org/posts/%0Ashow/7-testing>(0.1ms)
> > # commenting expires caches
> > Expired fragment: views/dev.pittsburghgardenexperiment.org/application/
> > footer_content (0.1ms)
> > Expired fragment: views/dev.pittsburghgardenexperiment.org/application/
> > footer_content (0.1ms)
> > Expired fragment: views/dev.pittsburghgardenexperiment.org/posts/show/
> > 7-testing<http://dev.pittsburghgardenexperiment.org/posts/show/%0A7-testing>(0.1ms)
> > Expired fragment: views/dev.pittsburghgardenexperiment.org/index
> > (0.1ms)
> > Expired fragment: views/dev.pittsburghgardenexperiment.org/categories/
> > 4-Talk
> > <http://dev.pittsburghgardenexperiment.org/categories/%0A4-Talk>(0.1ms)
> > # logout and refresh and we get a cache hit
> > Cached fragment hit: views/dev.pittsburghgardenexperiment.org/posts/
> > show/7-testing<http://dev.pittsburghgardenexperiment.org/posts/%0Ashow/7-testing>(0.1ms)
> > # refresh again and we get a hit, followed by a miss in the same
> > request. The correct content is rendered.
> > Cached fragment hit: views/dev.pittsburghgardenexperiment.org/posts/
> > show/7-testing<http://dev.pittsburghgardenexperiment.org/posts/%0Ashow/7-testing>(0.1ms)
> > Cached fragment miss: views/dev.pittsburghgardenexperiment.org/posts/
> > show/7-testing<http://dev.pittsburghgardenexperiment.org/posts/%0Ashow/7-testing>(0.1ms)
>
> > Any thoughts on why I'm getting stale data on the first request? Or
> > why the second request has a hit, followed by a miss?
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"CommunityEngine" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/communityengine?hl=en
-~----------~----~----~----~------~----~------~--~---
