Not entirely sure why it's sometimes doing it right, and sometimes  
doing it wrong, but it looks like you've got Sphinx running - and if  
that's the case, the indexer task assumes that every index it  
generates, it's replacing an existing one.

Try stopping Sphinx, deleting those files, re-indexing, then  
restarting Sphinx. And don't forget to stop Sphinx if you're ever  
adding or removing an index.

-- 
Pat

On 12/04/2009, at 12:00 PM, migrating_from_ultrasphinx wrote:

>
> Let me preface this by first stating that I have an application that
> has user data in separate database.  Each user has his own database.
> So I have hacked TS in it's configure and search code to generate
> index configs for each user.
>
> Having said that, running the ts:in task, which apparently just runs
> config and the Sphinx indexer, I get index files that look like this:
>
>
> ** Note, that "14014" is our user's account number, which specifies
> the db being indexed.
>
> -rw-r--r--   1 root  admin       791 Apr 11 21:40
> 14014_product_core.new.spp
> -rw-r--r--   1 root  admin      4740 Apr 11 21:40
> 14014_product_core.new.spm
> -rw-r--r--   1 root  admin      2709 Apr 11 21:40
> 14014_product_core.new.spi
> -rw-r--r--   1 root  admin       325 Apr 11 21:40
> 14014_product_core.new.sph
> -rw-r--r--   1 root  admin      2673 Apr 11 21:40
> 14014_product_core.new.spd
> -rw-r--r--   1 root  admin     12640 Apr 11 21:40
> 14014_product_core.new.spa
>
> Where is that ".new" coming from?
>
> And of course when I run ts:start, it complains that it can't find the
> indexes in the config file.  Here's the config:
>
> source 14014_product_core_0
> {
>  type = mysql
>  sql_host = localhost
>  sql_user = root
>  sql_pass = iware1pr
>  sql_db = c14014
>  sql_port = 3306
>  sql_query_pre = SET NAMES utf8
>  sql_query = SELECT `products`.`id` * 5 + 3 AS `id` , CAST
> (`products`.`part_number` AS CHAR) AS `part_number`, IFNULL
> (`products`.`part_number`, '') AS `part_number_sort`, UNIX_TIMESTAMP
> (`products`.`created_at`) AS `created_at`, UNIX_TIMESTAMP
> (`products`.`updated_at`) AS `updated_at`, `products`.`id` AS
> `sphinx_internal_id`, 485965105 AS `class_crc`, '485965105' AS
> `subclass_crcs`, 0 AS `sphinx_deleted` FROM products    WHERE
> `products`.`id` >= $start   AND `products`.`id` <= $end    GROUP BY
> `products`.`id`  ORDER BY NULL
>  sql_query_range = SELECT IFNULL(MIN(`id`), 1), IFNULL(MAX(`id`), 1)
> FROM `products`
>  sql_attr_uint = sphinx_internal_id
>  sql_attr_uint = class_crc
>  sql_attr_uint = sphinx_deleted
>  sql_attr_timestamp = created_at
>  sql_attr_timestamp = updated_at
>  sql_attr_str2ordinal = part_number_sort
>  sql_attr_multi = uint subclass_crcs from field
>  sql_query_info = SELECT * FROM `products` WHERE  `id` = (($id - 3) /
> 5)
> }
>
> index 14014_product_core
> {
>  source = 14014_product_core_0
>  path = /home/sphinx/sphinx/development/14014_product_core
>  morphology = stem_en
>  charset_type = utf-8
> }
>
> index 14014_product
> {
>  type = distributed
>  local = 14014_product_core
> }
>
> Here's the crazy thing though.  If I run config for just ONE account,
> the index file names are generated correctly.
> >


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