jackrabbit-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jukka Zitting" <jukka.zitt...@gmail.com>
Subject Re: Is doc addition / indexing synchronous or asynchronous?
Date Thu, 06 Apr 2006 17:50:33 GMT
Hi,

On 4/6/06, thomasg <thomasgascoigne@hotmail.com> wrote:
> Don't think I'm using transactions. How do you use transactions? (brief
> answer ok)

There's a XASession superinterface that allows you to use JTA
distributed transactions with a JCR repository like Jackrabbit.

> The 1 min for 50MB was for the whole process of adding, text extraction,
> indexing and running a query. I've retimed the call to session.save() as
> 9937ms (dropping to 2593ms when filters removed). The call to
> query.execute() is timed at 38859ms! Guess it must have to wait for the
> indexing to complete before it runs?

OK, thanks for the details. Are you receiving just a single or a few
results from the query or are there many results? It is a known
performance issue that by default the XPath query engine orders query
results in the "document order", which causes considerable performance
overhead for large result sets. You can try disabling this by adding
the following parameter to your SearchIndex configuration:

    <param name="respectDocumentOrder" value="false"/>

The XMLPersistenceManager is also not very fast, so you may want to
consider switching to for example DerbyPersistenceManager for better
performance.

BR,

Jukka Zitting

--
Yukatan - http://yukatan.fi/ - info@yukatan.fi
Software craftsmanship, JCR consulting, and Java development
Mime
View raw message