if you use the mysql command-line tool to connect to the test  
database, how quickly does the UPDATE statement (the first  
sql_query_pre) run?

-- 
Pat

On 16/10/2009, at 8:32 AM, rejeep wrote:

>
> Under *source klass_core_0*
> sql_query_pre = UPDATE `java_classes` SET `delta` = 0 WHERE `delta` =
> 1
> sql_query_pre = SET NAMES utf8
>
> And under *source klass_delta_0 : klass_core_0*
> sql_query_pre =
> sql_query_pre = SET NAMES utf8
>
> On Oct 15, 9:20 pm, Pat Allan <[email protected]> wrote:
>> Sorry, I should have been more clear... in your test.sphinx.conf  
>> file,
>> what are the sql_query_pre values in your klass_core_i source?
>>
>> --
>> Pat
>>
>> On 15/10/2009, at 11:16 PM, Pat Allan wrote:
>>
>>
>>
>>> Wow, something's really going wrong there with the sql_query_pre SQL
>>> statements... can you copy and paste the ones from the klass_core_1
>>> source?
>>
>>> --
>>> Pat
>>
>>> On 15/10/2009, at 7:51 PM, rejeep wrote:
>>
>>>> Sorry for the late reply.
>>
>>>> Here's the log:
>>>> Sphinx 0.9.9-rc2 (r1785)
>>>> Copyright (c) 2001-2009, Andrew Aksyonoff
>>
>>>> using config file '/home/rejeep/dev/closing/config/
>>>> test.sphinx.conf'...
>>>> indexing index 'klass_core'...
>>>> ERROR: index 'klass_core': sql_query_pre[0]: Lock wait timeout
>>>> exceeded; try restarting transaction (DSN=mysql://
>>>> closing:*...@localhost:3306/closing_test).
>>>> total 0 docs, 0 bytes
>>>> total 51.825 sec, 0 bytes/sec, 0.00 docs/sec
>>>> indexing index 'klass_delta'...
>>>> collected 0 docs, 0.0 MB
>>>> collected 0 attr values
>>>> sorted 0.0 Mvalues, nan% done
>>>> total 0 docs, 0 bytes
>>>> total 0.005 sec, 0 bytes/sec, 0.00 docs/sec
>>>> distributed index 'klass' can not be directly indexed; skipping.
>>>> indexing index 'interface_core'...
>>>> ERROR: index 'interface_core': sql_query_pre[0]: Lock wait timeout
>>>> exceeded; try restarting transaction (DSN=mysql://
>>>> closing:*...@localhost:3306/closing_test).
>>>> total 0 docs, 0 bytes
>>>> total 50.052 sec, 0 bytes/sec, 0.00 docs/sec
>>>> indexing index 'interface_delta'...
>>>> collected 0 docs, 0.0 MB
>>>> collected 0 attr values
>>>> sorted 0.0 Mvalues, nan% done
>>>> total 0 docs, 0 bytes
>>>> total 0.005 sec, 0 bytes/sec, 0.00 docs/sec
>>>> distributed index 'interface' can not be directly indexed;  
>>>> skipping.
>>>> indexing index 'enum_core'...
>>>> ERROR: index 'enum_core': sql_query_pre[0]: Lock wait timeout
>>>> exceeded; try restarting transaction (DSN=mysql://
>>>> closing:*...@localhost:3306/closing_test).
>>>> total 0 docs, 0 bytes
>>>> total 51.055 sec, 0 bytes/sec, 0.00 docs/sec
>>>> indexing index 'enum_delta'...
>>>> collected 0 docs, 0.0 MB
>>>> collected 0 attr values
>>>> sorted 0.0 Mvalues, nan% done
>>>> total 0 docs, 0 bytes
>>>> total 0.005 sec, 0 bytes/sec, 0.00 docs/sec
>>>> distributed index 'enum' can not be directly indexed; skipping.
>>>> indexing index 'annotation_core'...
>>>> ERROR: index 'annotation_core': sql_query_pre[0]: Lock wait timeout
>>>> exceeded; try restarting transaction (DSN=mysql://
>>>> closing:*...@localhost:3306/closing_test).
>>>> total 0 docs, 0 bytes
>>>> total 51.058 sec, 0 bytes/sec, 0.00 docs/sec
>>>> indexing index 'annotation_delta'...
>>>> collected 0 docs, 0.0 MB
>>>> collected 0 attr values
>>>> sorted 0.0 Mvalues, nan% done
>>>> total 0 docs, 0 bytes
>>>> total 0.005 sec, 0 bytes/sec, 0.00 docs/sec
>>>> distributed index 'annotation' can not be directly indexed;  
>>>> skipping.
>>>> total 4 reads, 0.000 sec, 32.0 kb/call avg, 0.0 msec/call avg
>>>> total 16 writes, 0.000 sec, 0.0 kb/call avg, 0.0 msec/call avg
>>>> rotating indices: succesfully sent SIGHUP to searchd (pid=11305).
>>>> Loaded suite /usr/bin/rake
>>
>>>> On Oct 11, 8:44 pm, Pat Allan <[email protected]> wrote:
>>>>> Right. Can you show me the output when it's running slowly?
>>
>>>>> --  
>>>>> Pat
>>
>>>>> On 11/10/2009, at 7:09 PM, rejeep wrote:
>>
>>>>>> That is not slow at all. But that's because the database does not
>>>>>> contain any records. But if I create a record from a factory in  
>>>>>> the
>>>>>> test and then run the rake task, also from the test, it will take
>>>>>> forever.
>>
>>>>>> On Oct 11, 1:36 pm, Pat Allan <[email protected]> wrote:
>>>>>>> Either I'm missing it, or there isn't actually any data being
>>>>>>> indexed
>>>>>>> (nor does it seem to be running slowly). What's the output (and
>>>>>>> how
>>>>>>> are you running it) when it *is* running slowly?
>>
>>>>>>> --
>>>>>>> Pat
>>
>>>>>>> On 10/10/2009, at 1:28 PM, rejeep wrote:
>>
>>>>>>>> The test database is exactly the same as the development. I  
>>>>>>>> don't
>>>>>>>> know
>>>>>>>> if it matters, but it's a cucumber test, so it's actually the
>>>>>>>> cucumber
>>>>>>>> environment. But that environment is only linking to test, so  
>>>>>>>> it
>>>>>>>> should be exactly the same.
>>
>>>>>>>> $ rake ts:in RAILS_ENV=test
>>>>>>>> (in /home/rejeep/dev/closing)
>>>>>>>> Generating Configuration to /home/rejeep/dev/closing/config/
>>>>>>>> test.sphinx.conf
>>>>>>>> Sphinx 0.9.9-rc2 (r1785)
>>>>>>>> Copyright (c) 2001-2009, Andrew Aksyonoff
>>
>>>>>>>> using config file '/home/rejeep/dev/closing/config/
>>>>>>>> test.sphinx.conf'...
>>>>>>>> indexing index 'klass_core'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.005 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> indexing index 'klass_delta'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.001 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> distributed index 'klass' can not be directly indexed;  
>>>>>>>> skipping.
>>>>>>>> indexing index 'interface_core'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.002 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> indexing index 'interface_delta'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.001 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> distributed index 'interface' can not be directly indexed;
>>>>>>>> skipping.
>>>>>>>> indexing index 'enum_core'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.001 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> indexing index 'enum_delta'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.001 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> distributed index 'enum' can not be directly indexed; skipping.
>>>>>>>> indexing index 'annotation_core'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.001 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> indexing index 'annotation_delta'...
>>>>>>>> collected 0 docs, 0.0 MB
>>>>>>>> collected 0 attr values
>>>>>>>> sorted 0.0 Mvalues, nan% done
>>>>>>>> total 0 docs, 0 bytes
>>>>>>>> total 0.002 sec, 0 bytes/sec, 0.00 docs/sec
>>>>>>>> distributed index 'annotation' can not be directly indexed;
>>>>>>>> skipping.
>>>>>>>> total 8 reads, 0.000 sec, 32.0 kb/call avg, 0.0 msec/call avg
>>>>>>>> total 32 writes, 0.001 sec, 0.0 kb/call avg, 0.0 msec/call avg
>>>>>>>> Loaded suite /usr/bin/rake
>>>>>>>> Started
>>
>>>>>>>> Finished in 0.000242 seconds.
>>
>>>>>>>> 0 tests, 0 assertions, 0 failures, 0 errors
>>
>>>>>>>> On Oct 10, 10:42 am, Pat Allan <[email protected]>  
>>>>>>>> wrote:
>>>>>>>>> If there's only one record, then it shouldn't matter what your
>>>>>>>>> range
>>>>>>>>> step is, because Sphinx finds the smallest and largest ids
>>>>>>>>> available -
>>>>>>>>> which in this case, are the same number.
>>
>>>>>>>>> Do you have all the same database indexes on your test db and
>>>>>>>>> development db? What's the output of rake ts:in  
>>>>>>>>> RAILS_ENV=test?
>>
>>>>>>>>> --
>>>>>>>>> Pat
>>
>>>>>>>>> On 09/10/2009, at 8:00 PM, rejeep wrote:
>>
>>>>>>>>>> I didn't mean that I was trying to index a single record.
>>>>>>>>>> What I
>>>>>>>>>> meant
>>>>>>>>>> was that I only had one record in the database for that test.
>>>>>>>>>> So
>>>>>>>>>> even
>>>>>>>>>> if there would be some high index, I guess it wouldn't take  
>>>>>>>>>> to
>>>>>>>>>> long
>>>>>>>>>> anyway.
>>
>>>>>>>>>> If I try with this conf:
>>>>>>>>>> development:
>>>>>>>>>> sql_range_step: 1000
>>>>>>>>>> production:
>>>>>>>>>> sql_range_step: 1000
>>>>>>>>>> test:
>>>>>>>>>> sql_range_step: 1000
>>
>>>>>>>>>> It's fast in development, but equally slow in test.
>>
>>>>>>>>>> It doesn't matter if I use 1000 or 10000000 for  
>>>>>>>>>> sql_range_step.
>>
>>>>>>>>>> On Oct 9, 5:43 pm, Pat Allan <[email protected]>  
>>>>>>>>>> wrote:
>>>>>>>>>>> Hi Rejeep
>>
>>>>>>>>>>> Sphinx doesn't allow you to process a single record - only a
>>>>>>>>>>> full
>>>>>>>>>>> index. Also, do you have sql_range_step set for the test
>>>>>>>>>>> environment,
>>>>>>>>>>> as well as the development environment?
>>
>>>>>>>>>>> --
>>>>>>>>>>> Pat
>>
>>>>>>>>>>> On 09/10/2009, at 11:53 AM, rejeep wrote:
>>
>>>>>>>>>>>> Hi!
>>
>>>>>>>>>>>> I have a site where I want to test the search. From the
>>>>>>>>>>>> test I
>>>>>>>>>>>> create
>>>>>>>>>>>> the records and then create the index. But it's so slow  
>>>>>>>>>>>> that
>>>>>>>>>>>> it
>>>>>>>>>>>> times
>>>>>>>>>>>> out. Yes, I'm using factories and I am aware of the id
>>>>>>>>>>>> problem.
>>>>>>>>>>>> But
>>>>>>>>>>>> first of all. In the test I only want to index 1 record.  
>>>>>>>>>>>> And
>>>>>>>>>>>> the
>>>>>>>>>>>> id of
>>>>>>>>>>>> that usually is between 1000 - 5000. So that should not be
>>>>>>>>>>>> such a
>>>>>>>>>>>> big
>>>>>>>>>>>> problem, right? And even thought I set sql_range_step, it  
>>>>>>>>>>>> is
>>>>>>>>>>>> still
>>>>>>>>>>>> slow.
>>
>>>>>>>>>>>> Since it is super fast in development the only thing I  
>>>>>>>>>>>> could
>>>>>>>>>>>> think
>>>>>>>>>>>> be
>>>>>>>>>>>> the problem was the id's. But since it's not. Does someone
>>>>>>>>>>>> has
>>>>>>>>>>>> any
>>>>>>>>>>>> other suggestion?
>>
>>>>>>>>>>>> If I skip creating any records before the indexing in the
>>>>>>>>>>>> test.
>>>>>>>>>>>> Then
>>>>>>>>>>>> the indexing is fast.
>>
>>
> >


--~--~---------~--~----~------------~-------~--~----~
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