[EP-tech] Re: IRSTATS2 questions

[Yuri]

Shouldn't early result be cached, and only new hits be counted?

Il 05/09/2013 20:19, Betsy Coles ha scritto:
>
> Hello Séb et al.,
>
> I have IRSTATS2-0.0.4 beta from early July installed on my main 
> repository. I’m very impressed – it’s exactly what we need – but I 
> have three questions:
>
> 1.Is there a newer version? How can I get it?
>
> 2.There are some issues with capitalization of names in the authors 
> lists, e.g., middle initials are lower case, as “Buchwald, Jed z.” 
> This looks like it should be simple to patch but I’m not sure where to 
> do so. Hints?
>
> 3.Performance is definitely an issue. My repository has 35,000 items 
> and a 10-year history resulting in an access table with 79,000,000 
> rows. IRSTATS2 reports almost 4,000,000 downloads. The initial 
> process_stats job took 40 hours to run on a fairly capable machine. 
> Worse, the “5 Top Authors” list on the summary page takes 45 seconds 
> to load. And the table of countries (not map) in the Requests report 
> is still trying to load after half an hour of spinning. Are there 
> perhaps plans to improve performance?
>
> You can see our stats at 
> http://authors.library.caltech.edu/cgi/stats/report 
> <http://authors.library.caltech.edu/cgi/stats/report> (but don’t 
> everybody hit it at once, please!).
>
> Many thanks,
>
> Betsy Coles
>
> Caltech Library IT Group
>
> bcoles at caltech.edu
>
>
>
> *** Options: http://mailman.ecs.soton.ac.uk/mailman/listinfo/eprints-tech
> *** Archive: http://www.eprints.org/tech.php/
> *** EPrints community wiki: http://wiki.eprints.org/