Re: Question about exist:match

On our server running rev8536 we get errors such as

[Fatal Error] :2:113: The element type "model" must be terminated by the 
matching end-tag "</model>".
[Fatal Error] :2:101: The element type "model" must be terminated by the 
matching end-tag "</model>".

appearing on the command line where exist was started.  How do I force 
all errors to be logged or reported to users? Or is it impossible due to 
external libraries?

This email was independently scanned for viruses by McAfee anti-virus software and none were found
------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

memory problem

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open

Re: memory problem

 > I got as far as the opening screen of the admin section, before I got 
the out
> of memory error again.

So you say you did not even store any data into the db? This can't be an 
eXist problem. Any messages in the logs?

Wolfgang

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

Windows Server 2008 / Xeon multicore

Hi,

what can I do if my eXist xqueries run slower on the new server (old: Windows XP, 1.5G RAM, single
core; new: Windows Server 2008, 4G RAM, 4 Xeon cores) where java uses only 25% of processor cycles?
I guess it's just using a single core? What are my options here?

Thanks, wuuf

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

Re: Windows Server 2008 / Xeon multicore

Hi,

> what can I do if my eXist xqueries run slower on the new server (old: Windows XP, 1.5G RAM, single
> core; new: Windows Server 2008, 4G RAM, 4 Xeon cores) where java uses only 25% of processor cycles?
> I guess it's just using a single core? What are my options here?

My experience on Win XP/Vista is that java automatically distributes 
threads on the available cores, so no further configuration should be 
required here. Do you use the same eXist settings (cacheSize, memory 
available to Java) on both machines?

Wolfgang

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

Re: Windows Server 2008 / Xeon multicore

> From: wuuf <at> gmx.net [wuuf <at> gmx.net]
> what can I do if my eXist xqueries run slower on the new server
> (old: Windows XP, 1.5G RAM, single core; new: Windows Server
> 2008, 4G RAM, 4 Xeon cores) where java uses only 25%
> of processor cycles? I guess it's just using a single core?
> What are my options here?

What's the processor clock, memory and disk speeds on the new box compared to the old one?  Is it the same eXist
version on both?  Which version?

Are you running multiple concurrent queries, or just single ones at a time?  Are you benchmarking with a
smaller load than your real load, or is this happening in your production environment?

Wolfgang and other committers, does eXist make any use of parallelism to evaluate a single query, or is one
query evaluated on one thread?  I know it's a *very* hard job to get parallel query evaluation right!

- Peter

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

Two questions about the xmlrpc API

Hi, I got two questions about the xmlrpc API. I'm using python.

1: I'm using parseLocal(String localFile, String docName, boolean 
replace, String mimeType) to store xml files in eXist. It works, but it 
has the annoying side effect of removing the file from my computer, why? 
I guess I could open it and read it myself and call parse(String xml, 
String document, int overwrite) to get around this problem, but that 
feels inconvenient. Why on earth does eXist think that I want to remove 
the file passed to parseLocal()?

2: parseLocal(), takes a boolean deciding if an eventual exisiting 
resource should be written or not. Very good. But parse(String xml, 
String document, int overwrite) takes an int instead. Why can't they 
both take a boolean?

- EL

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

Re: Windows Server 2008 / Xeon multicore

> Wolfgang and other committers, does eXist make any use of parallelism to evaluate a single query, or is one
query evaluated on one thread?

Each query will run in its own thread. We could probably make some use 
of parallelism by 1) providing a threading XQuery module, 2) splitting 
cpu intensive index lookup or node set merge operations into separate 
threads. However, concerning 2), I believe that further improvements to 
query optimization are more promising. There's still a lot of potential 
in there.

Wolfgang

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

Re: Windows Server 2008 / Xeon multicore

Wuff asked:

> what can I do if my eXist xqueries run slower on the new server (old: 
> Windows XP, 1.5G RAM, single
> core; new: Windows Server 2008, 4G RAM, 4 Xeon cores) where java uses 
> only 25% of processor cycles?
> I guess it's just using a single core? What are my options here?

Convert to Linux? 

--

-- 
Andrzej Taramina
Chaeron Corporation: Enterprise System Solutions
http://www.chaeron.com

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H

Predicates scope problem?

Hello,

I find something strange in Exist stable (with Saxon 9b or Xalan).

In xpath, I know that "//statut[1]" is different of
"/descendant-or-self::statut[1]".

with this document...

<articles>
    <article>
        <statuts>
            <statut no="1"/>
            <statut no="2"/>
            <statut no="3"/>
        </statuts>
    </article>
    <article>
        <statuts>
            <statut no="10"/>
            <statut no="11"/>
        </statuts>
    </article>
</articles>

with "/descendant-or-self::statut[1]", I must receive only "<statut
no="1"/>" (the fisrt <statut> of the document).

But with "//statut[1]", I must receive "<statut no="1"/><statut
no="10"/>" (the first <statut> of all <statuts>).

First problem : Exist translates "//statut[1]" to
"/descendant-or-self::statut[1]" wich doesn't mean the same thing.
It should be "/descendant-or-self::element()/statut[1]".

So I search "/descendant-or-self::element()/statut[1]".
Exist translates to "/descendant-or-self::element()/child::statut[1]"
(it's ok for me).

1) xpath : /descendant-or-self::element()/statut[1]
results :
<statut no="1"/>
<statut no="2"/>
<statut no="3"/>
<statut no="10"/>
<statut no="11"/>

2) xpath : /descendant-or-self::element()/statut[2]
results :
<statut no="2"/>
<statut no="11"/>

3) xpath : /descendant-or-self::element()/statut[3]
results :
<statut no="3"/>

4) xpath : /descendant-or-self::element()/statut[4]
results :
<statut no="10"/>

2 and 3 are correct
1 and 4 are incorrect

1 should give <statut no="1"/><statut no="10"/>
4 should give nothing

Am I wrong?

Thanks

Dominic

------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H