Mailing-List: contact users-help@jackrabbit.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@jackrabbit.apache.org
Received-SPF: pass (athena.apache.org: local policy includes SPF record at
 spf.trusted-forwarder.org)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=yahoo.co.uk;
  h=Received:X-YMail-OSG:X-Yahoo-Newman-Property:From:To:References:In-Reply-To:Subject:Date:Message-ID:MIME-Version:Content-Type:Content-Transfer-Encoding:X-Mailer:Thread-Index:Content-Language;
  b=okPURbZIbPi0j0YBPAVXbPJZOTJGkZtFy7t7aGUbWsfElhO62A+HGRvEK9C/GvWI85zLUsbaGIrHPwZ2HdgCD5v9IeXS1lNpR419BifYMkc9Ljvmm9fLkSOMVI7P9YBn6aFGdqILXPoauI1OX23maVcdwFv/ChYDdliseHpiw+w=
  ;
From: "sbarriba" <sbarriba@yahoo.co.uk>
To: <users@jackrabbit.apache.org>
References: <08b601c8d5ce$8aed0eb0$a0c72c10$@co.uk> <4860C270.3050700@gmx.net>
In-Reply-To: <4860C270.3050700@gmx.net>
Subject: RE: Logging/Detecting slow JCR queries with Lucene
Date: Tue, 24 Jun 2008 11:20:29 +0100
Message-ID: <097201c8d5e3$f1dc7ab0$d5957010$@co.uk>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
Thread-Index: AcjV30J9ypzJIczuTtO/LoXssL6X3AABHM7w
Content-Language: en-gb

Thanks Marcel. We'll look at enabling this logging.

Are there any tuning params we should be looking at to help with the
performance in this area in the short term?

Regards,
Shaun


-----Original Message-----
From: Marcel Reutegger [mailto:marcel.reutegger@gmx.net] 
Sent: 24 June 2008 10:46
To: users@jackrabbit.apache.org
Subject: Re: Logging/Detecting slow JCR queries with Lucene

Hi,

sbarriba wrote:
> Hi all,
> 
> Our application makes quite heavy use of JCR queries. We're seeing some
> scenarios where the IO utilisation is very high for a period of time. A
> thread dump of Tomcat shows lots and lots of threads locked waiting on
> <0x781fe1d0> (a org.apache.lucene.store.FSDirectory$FSIndexInput which is
> locked by the thread below.

not sure, but this might be related to: 
https://issues.apache.org/jira/browse/JCR-1365

> Given we have 140+ queries being used, is there a way to configure
> JackRabbit and/or Lucene to log out slow queries equivalent to the
> "log-slow-queries" provided by MySql?

when you enable debug log messages for 
org.apache.jackrabbit.core.query.QueryImpl you will see timing information
as 
well as the query statement, though you'll have to post process that
information 
because *all* queries will be logged, not just slow ones.


regards
  marcel