Re: [Exist-development] Index Configuration // Performance Issue

[<Lar...@ie...>]

Hi Leif-Jöran,

the question from 2010-12-20 is solved.
I sent Wolfgang some XML-Data and he adviced me to use the new eXist
1.5 because also an optimised query was processed to slow with eXist 1.4
In this mail I refer to an an other query which seems to be not optimal
cause the eXist-Log has
> TRACE (Optimize.java [getQNameIndexType]:291) - Cannot optimize:
collection /db/loewe-a3/tocs/layers/txsf_u8 does not define an index on
start
> TRACE (Optimize.java [eval]:169) - exist:optimize: Cannot optimize
expression.

and the Query is processed way to slow, compared to the
Saxon-Processor.

Thanks for your help

LOK

 --- Ursprüngliche Nachricht ---
Datum: 31.01.2011 16:49
Von: Leif-Jöran Olsson <lj...@ex...>
An: Lars-Olof Krause <Lar...@ie...>
Betreff: Re: [Exist-development] Index Configuration // Performance
Issue

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Dear Lars-Olof, I think the majority of the questions originating
from
> your 2010-12-20 letter on the subject of Performance problem are
still
> without answer from you. I have not seen any xml snippet so I suspect
> you only sent it to Wolfgang?
>
> Cheers,
> Leif-Jöran
>
> Den 2011-01-31 15:50, Lars-Olof Krause skrev:
> > Hi,
> >
> > I write the the development-list cause I got no answer from the
> > open-list and the question is somehow related to the development
cause
> > it deals with performance issues.
> >
> > I got an quite complex xQuery which processes the whole collection
> > using an Saxon-Processor in several seconds.
> > Using my eXist1.5 the process of one Dokument takes way longer than
the
> > processing of the whole collection using the saxon-processor.
> >
> > Since my eXist-Log shows
> > TRACE (Optimize.java [getQNameIndexType]:291) - Cannot optimize:
> > collection /db/loewe-a3/tocs/layers/txsf_u8 does not define an
index on
> > start
> > TRACE (Optimize.java [eval]:169) - exist:optimize: Cannot optimize
> > expression.
> >
> > Something has to be wrong with my indexes.
> >
> > Can someone tell me which indexes I should configure for the
following
> > xquery-script or which changes I could do to improve the
> > script-performance?
> >
> > Thanks so much.
> >
> > LOK
> >
> > xQuery:
> > ----------------------------------------
> > xquery version "1.0";
> >
> > declare boundary-space strip;
> >
> > declare namespace
xsf="http://www.xstandoff.net/2009/xstandoff/1.1";
> > declare namespace xsi="http://www.w3.org/2001/XMLSchema-instance";
> > declare namespace tocs="http://taurus.zmi.uni-giessen.de/tocs";
> > declare namespace terms="http://taurus.zmi.uni-giessen.de/terms";
> > declare namespace cnxde="http://www.text-technology.de/cnxde";
> > declare namespace my="http://www.taurus.zmi.uni-giessen.de/my";
> > declare default element namespace
> > "http://www.xstandoff.net/2009/xstandoff/1.1";
> >
> > declare function my:primdata($node as node(), $start as xs:integer,
> > $end as xs:integer) as xs:string {
> >     substring($node, $start + 1 , $end - $start)
> > };
> >
> > declare function my:term2lemmaRep($termSegment as node(), $d as
node())
> > as xs:string {
> >    let $token := $d//cnxde:token[./id(@xsf:segment)/@start+0 ge
> > $termSegment/@start+0][./id(@xsf:segment)/@end+0 le
> > $termSegment/@end+0]
> >    return string(($token[1]/@lemma))
> > };
> >
> > <result-document>
> > {
> >   let $doc :=
> >
doc("loewe-a3/tocs/layers/txsf_u8/raithel_paedagogik_2007-txsf.xml")
> >       let $text := $doc//xsf:textualContent
> >
> >     return
> >     for $child in $doc//tocs:label
> >
> >         let $childSegment :=  $child/id(@xsf:segment)
> >         let $childRange := my:primdata($text, $childSegment/@start,
> > $childSegment/@end)
> >
> >         let $childTerm :=
($doc//terms:term/id(@xsf:segment)[@start+0
> > ge $childSegment/@start+0][@end+0 le $childSegment/@end+0])[1]
> >
> >         where $childTerm
> >         return
> >         let $childTermRange := my:primdata($text,
$childTerm/@start,
> > $childTerm/@end)
> >         let $childTermLemmaRep := my:term2lemmaRep($childTerm,$doc)
> >
> >         let $self :=
> > $child/../../../preceding-sibling::tocs:item[1]/tocs:label[1]
> >
> >         let $selfSegment := $self/id(@xsf:segment)
> >         let $selfRange := my:primdata($text, $selfSegment/@start,
> > $selfSegment/@end)
> >
> >         where $self and ( contains($selfRange, "Disziplinen") or
> >                           contains($selfRange, "Subdisziplinen") or
> >                           contains($selfRange, "Fachrichtungen") or
> >                           contains($selfRange, "Hauptrichtungen")
or
> >                           contains($selfRange, "gebiete") or
> >                           contains($selfRange, "fächer") or
> >                           contains($selfRange, "Bereiche") or
> >                           contains($selfRange, "Zweige") )
> >
> >
> >         return
> >         let $parent :=
> > $self/../../../preceding-sibling::tocs:item[1]/tocs:label[1]
> >         let $parentSegment := $parent/id(@xsf:segment)
> >         let $parentRange := my:primdata($text,
$parentSegment/@start,
> > $parentSegment/@end)
> >
> >         let $parentTerm :=
($doc//terms:term/id(@xsf:segment)[@start ge
> > $parentSegment/@start][@end le $parentSegment/@end])[1]
> >
> >         where $parentTerm
> >         return
> >         let $parentTermRange := my:primdata($text,
$parentTerm/@start,
> > $parentTerm/@end)
> >         let $termLemmaRep := my:term2lemmaRep($parentTerm, $doc)
> >         return
> >         <construction>
> >             <parent>
> >                <label>{$parentRange}</label>
> >                <term>{$parentTermRange}</term>
> >                <termLemmaRep>{$termLemmaRep}</termLemmaRep>
> >
> >             </parent>
> >             <self>
> >                 <label>{$selfRange}</label>
> >             </self>
> >             <child>
> >                <label>{$childRange}</label>
> >                <term>{$childTermRange}</term>
> >                <termLemmaRep>{$childTermLemmaRep}</termLemmaRep>
> >             </child>
> >         </construction>
> > }
> > </result-document>
> >
> >
> >
------------------------------------------------------------------------------
> > Special Offer-- Download ArcSight Logger for FREE (a $49 USD
value)!
> > Finally, a world-class log management solution at an even better
price-free!
> > Download using promo code Free_Logger_4_Dev2Dev. Offer expires
> > February 28th, so secure your free ArcSight Logger TODAY!
> > http://p.sf.net/sfu/arcsight-sfd2d
> > _______________________________________________
> > Exist-development mailing list
> > Exi...@li...
> > https://lists.sourceforge.net/lists/listinfo/exist-development
> >
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.16 (GNU/Linux)
> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
>
> iD8DBQFNRtn7hcIn5aVXOPIRAuG1AJ93y+rvVIH8AsZdJDlFf6+gGdqjxQCgsUgr
> 64BjQZkxg2jaq8jmKIgZxww=
> =kjxn
> -----END PGP SIGNATURE-----
>