Ah, that's an annoying flaw indeed. I'm glad you were able to figure it out... 
if you could log a ticket, then I'll try to reproduce it locally and make sure 
it doesn't happen again.

Sorry for the hassle.

-- 
Pat

On 04/03/2010, at 1:31 PM, Andrei wrote:

> Problem solved.
> 
> This happens if you configure your production to run in threadsafe
> mode.
> 
> On Mar 3, 2:04 pm, Andrei <[email protected]> wrote:
>> I have some progress.
>> 
>> The delta was not working because the TS was in development mode,
>> while the application was in production mode.
>> TS was trying to access a non-existent production config which we
>> could not generate. This incorrect approach worked in the earlier
>> versions of the TS for unknown reasons though :)
>> 
>> And as for production-mode config:
>> 
>> Inside the configuration.rb's build method the
>> ThinkingSphinx.context.indexed_models is always empty in production
>> mode.
>> 
>> ThinkingSphinx.context initializes itself with these two methods:
>> load_models and add_indexed_models in respective order.
>> 
>> The .constantize method in load_models always throws NameError
>> exception on every model in production mode, meaning there is not a
>> single model initialized at that moment. I'm not competent enough to
>> determine if this is the correct behavior or not, and what to do about
>> it.
>> And add_index_models, consequently (or not) cant see a single indexed
>> model.
>> 
>> I just cant understand, by looking at the code, why the development
>> mode works correctly and the production mode does not.
>> 
>> Can you please give some advice on where i can try to look next?
>> 
>> On Mar 2, 9:30 pm, Andrei <[email protected]> wrote:
>> 
>> 
>> 
>>> Our server have recently migrated to the passenger 2.2.9 and TS1.3.16,
>>> and i have noticed that the delta-indexing have stopped working since.
>>> Database is being updated with 1 in delta column, but the indexes just
>>> wont rotate, and the log is empty.
>> 
>>> Manually executing rake tasks works perfectly.
>> 
>>> We do not have a full control over our server. Just the superficial
>>> ssh user access. I do not know how to debug this problem from my end.
>>> TS just either works or not, it does not give any intermediate
>>> information. Is there a way to make TS more verbose?
>> 
>>> I have tried reading 'common questions and issues' and setting the
>>> 'bin_path', but the TS says that it cant find sphinx on my system no
>>> matter what path i provide (the path is correct). Before the problem
>>> everything worked without specifying the 'bin_path' though.
>> 
>>> And another thing: "rake ts:conf ENV_RAILS=production" always
>>> generates an almost empty config file for some reason — i can never
>>> run TS in production environment. Development environment works like
>>> expected.
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Thinking Sphinx" 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/thinking-sphinx?hl=en.
> 

-- 
You received this message because you are subscribed to the Google Groups 
"Thinking Sphinx" 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/thinking-sphinx?hl=en.

Reply via email to