Thanks for your emails Marcelo - they haven't been ignored, just trying to find some time to investigate.
What server set up are you using? Mongrel? Passenger (with Apache? Nginx?)? Something else? If you're open to giving me access to your codebase so I can test with that straight away, that'd be great, but not a requirement. -- Pat On 07/06/2010, at 6:41 AM, Marcelo Barbudas wrote: > At least in my setup it didn't work. > > After some investigations I found that in development mode slowdowns > started occurring in abe321df837f825db6bb. Previous version > 9f9f09a79e1e77b9cd6e worked perfectly. > > With the new version the first load is OK, afterwards @@context is > again nil and goes through re-loading all. > > -- > M. > > On Jun 6, 11:36 pm, def <[email protected]> wrote: >> Thanks for fast reply, I will play with that tomorrow :) >> hope it will helps, have nice day >> >> On 5 Чер, 08:24, Pat Allan <[email protected]> wrote: >> >> >> >>> Okay, just pushed 1.3.17 - you'll want to add something like the following >>> to the end of your environment.rb - before any references to any of your >>> models, but after the configure block: >> >>> ThinkingSphinx.indexed_models = ['ModelOne', 'ModelTwo'] >> >>> Having the model names as strings, not actual models, is important - if you >>> use actual models, then that evaluates the define_index block within the >>> model, which fires off TS's default handling of model loading. >> >>> Give it a go, hopefully it helps matters. >> >>> Also, for the record: there's no old versions that are fast at loading >>> models - it's important for TS to know about all models that are available >>> for searching. >> >>> -- >>> Pat >> >>> On 05/06/2010, at 4:37 AM, def wrote: >> >>>> Hi Pat, >> >>>> can you advise version that doesn't do this overhead? It become quite >>>> hard to work on development. >>>> I hope you can make this improvement in short time :) >> >>>> Thanks >>>> Def >> >>>> On Jun 3, 9:05 am, Pat Allan <[email protected]> wrote: >>>>> Hi Marcelo >> >>>>> At this point, there's no way to tell TS exactly which models to load. >>>>> You're not the first to request it, though, so I'll try to get something >>>>> into the code which does that. >> >>>>> -- >>>>> Pat >> >>>>> On 03/06/2010, at 6:46 AM, Marcelo Barbudas wrote: >> >>>>>> Hi Pat, >> >>>>>> Sorry to be answering so late. >> >>>>>> I am using the latest github version. >> >>>>>> This happens only in development mode right? Is there a way to >>>>>> explicitly tell TS what to load so I can benchmark without ts being a >>>>>> factor? >> >>>>>> Or to specify the models manually to avoid the extra CPU burn? >> >>>>>> -- >>>>>> M. >> >>>>>> On Jun 1, 10:31 am, Pat Allan <[email protected]> wrote: >>>>>>> Hi Marcelo >> >>>>>>> How many models do you have in your application? >> >>>>>>> To process search results, TS needs to know about *all* models that >>>>>>> have indexes, so it forces Rails to load every single model. >>>>>>> Unfortunately this can't be avoided, but it should only happen (with >>>>>>> the latest version of TS) when you're running a search. >> >>>>>>> So: what version of TS are you using? >> >>>>>>> -- >>>>>>> Pat >> >>>>>>> On 01/06/2010, at 5:23 AM, Marcelo Barbudas wrote: >> >>>>>>>> Hi. >> >>>>>>>> I've been profiling a Rails Application and noticed some big >>>>>>>> improvements once I disable TS. >> >>>>>>>> Basically if I have a model called Post which has some indexes. >>>>>>>> Once I remove define_index, in any method that uses Post, even in >>>>>>>> basic Post.find without any TS interaction, I get 4x improvements. >> >>>>>>>> I am using Rails 2.3.5 in development mode, testing via the NewRelic >>>>>>>> plugin. >> >>>>>>>> Any ideas? I don't even know what data to include. >> >>>>>>>> -- >>>>>>>> M. >> >>>>>>>> -- >>>>>>>> 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 >>>>>>>> athttp://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 >>>>>> athttp://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 >>>> athttp://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. > -- 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.
