it's generated normally - the ids range from 1 to 13966 with ~4000
items - although for the user index which the stats above are from
it's just 1 to 1538 with 1538 items.

the other two factors which I'm looking at are:
1. whilst the db box is fast and powerful, dual quad core processors /
15k SAS disks. the query is bottlenecking on being constrained to one
CPU core - the individual cores aren't THAT powerful and so could
quite conceivably be outdone by my laptop
2. database design, theres quite a lot of rails polymorphic columns,
which had the default 255 length and utf-8 collation = quite a lot of
bytes. I've changed this to varchar(14) with latin1_general_ci
collation. I didn't have a massive boost on my laptop - a few seconds
quicker, just running it on the server now...

cheers,

On Mar 12, 4:54 am, Pat Allan <[email protected]> wrote:
> Hi Nick
>
> How was your production data generated - normally, or by fixtures? (Or even: 
> what's the min and max ids in your user table?) If there's a really massive 
> window, then Sphinx is making a lot of queries to get through all the data, 
> but you can change 
> that:http://freelancing-god.github.com/ts/en/common_issues.html#slow_indexing
>
> If that doesn't help matters, let us know.
>
> Cheers
>
> --
> Pat
>
> On 12/03/2010, at 9:09 AM, Nick Sellen wrote:
>
>
>
> > I've got a case of very slow indexing - 111 seconds to index 9
> > documents. It's not really thinking sphinx that is responsible for
> > this part anyway - but just looking for any tips or ideas.
>
> > - on my laptop it runs fine, e.g.:
>
> > indexing index 'user_delta'...
> > total 2 docs, 68 bytes
> > total 0.048 sec, 1425.43 bytes/sec, 41.92 docs/sec
>
> > - on the production server one time it ran like this:
>
> > indexing index 'user_delta'..
> > total 9 docs, 320 bytes
> > total 111.667 sec, 2.87 bytes/sec, 0.08 docs/sec
>
> > - and another like this:
>
> > indexing index 'user_delta'...
> > total 14 docs, 533 bytes
> > total 8.589 sec, 62.06 bytes/sec, 1.63 docs/sec
>
> > the indexer is running on seperate box to the db - the db server is a
> > very fast database server (SAS disks etc) so the hardware should be
> > fine, but the load on it is surprisingly high when running the
> > indexing.
>
> > I ran a MySQL EXPLAIN command on the generated SQL configuration and
> > it shows all indexes being used.
>
> > so, yeah, any experiences people might have that could help would be
> > much appreciated.
>
> > cheers!
>
> > --
> > 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.

Reply via email to