score retrieval performance

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

score retrieval performance

Michael Ryan
Is there any significant difference in query speed when retrieving the score pseudo-field? E.g., does...

q=foo&sort=date+desc&fl=*,score

...take longer to run than...

q=foo&sort=date+desc&fl=*

I know there's different code paths in Solr depending on whether the score is needed or not, but not sure how much slower the score path is...

-Michael
Reply | Threaded
Open this post in threaded view
|

Re: score retrieval performance

Gora Mohanty-3
On 19 May 2014 19:52, Michael Ryan <[hidden email]> wrote:
> Is there any significant difference in query speed when retrieving the score pseudo-field? E.g., does...
>
> q=foo&sort=date+desc&fl=*,score
>
> ...take longer to run than...
>
> q=foo&sort=date+desc&fl=*
>
> I know there's different code paths in Solr depending on whether the score is needed or not, but not sure how much slower the score path is...

If you are using the q parameter, the score is being calculated in any
case, and while we
have never measured this, I would imagine that the performance hit
from retrieving the score
would be miniscule.

Are you perchance thinking of using filter queries using the fq
parameter? As filter queries are
cached, using these can indeed make a difference.

Regards,
Gora