Hi Anibal
That behaviour isn't what we want... I'll try to add some more tests,
see if I can reproduce the situation, and we'll go from there. I'm not
convinced with how effective Sphinx's merge is, hence why it's not yet
implemented with the basic delta approach.
Will let you know what I find. In the meantime, what does your
define_index block look like? Just as a reference point.
Cheers
--
Pat
On 16/01/2009, at 4:22 AM, anibal wrote:
>
> Hey,
>
> I'm trying to introduce the new datetime-driven deltas in my project
> but something very strange is happening. This is the situation:
>
> I create 3 objects 'Topic1', 'Topic2' and 'Topic3' and run a full
> index (rake ts:index). Then I create another object 'Topic4' and try
> searching all of them by their name. The sphinx query log reads like
> this:
>
> [Thu Jan 15 18:26:51.152 2009] 0.000 sec [all/2/rel 1 (0,20)] [*]
> Topic1
> [Thu Jan 15 18:26:54.239 2009] 0.001 sec [all/2/rel 1 (0,20)] [*]
> Topic2
> [Thu Jan 15 18:26:56.320 2009] 0.001 sec [all/2/rel 1 (0,20)] [*]
> Topic3
> [Thu Jan 15 18:27:02.144 2009] 0.000 sec [all/2/rel 0 (0,20)] [*]
> Topic4
>
> Which is perfect because 'Topic1', 'Topic2' and 'Topic3' are indexed
> but 'Topic4' is not.
>
> Then I run 'thinking_sphinx:index:delta' to merge the latest changes
> and try searching all of them again. The query log reads like this:
>
> [Thu Jan 15 18:28:10.400 2009] 0.000 sec [all/2/rel 0 (0,20)] [*]
> Topic1
> [Thu Jan 15 18:28:12.848 2009] 0.000 sec [all/2/rel 0 (0,20)] [*]
> Topic2
> [Thu Jan 15 18:28:14.839 2009] 0.000 sec [all/2/rel 0 (0,20)] [*]
> Topic3
> [Thu Jan 15 18:28:17.024 2009] 0.000 sec [all/2/rel 1 (0,20)] [*]
> Topic4
>
> Basically, only the new object 'Topic4' just merged is found but none
> of the previous objects are found.
>
> Just to add more information, if I run an empty query:
> ThinkingSphinx::Search.search(''), sphinx finds all of them (the 4
> topics):
>
> [Thu Jan 15 18:28:38.544 2009] 0.000 sec [scan/2/rel 4 (0,20)] [*]
>
> I was using the old delta mechanism and everything worked fine but I
> really wanna start using the new approach because is a lot better in
> my opinion.
>
> Hope someone could help me and thanks in advance,
> Anibal
>
> >
--~--~---------~--~----~------------~-------~--~----~
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
-~----------~----~----~----~------~----~------~--~---
