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)
Message-Id: <FF6CEC4D-CA10-40E5-B1F1-860B92ABE6F1@sptci.com>
From: Rakesh Vidyadharan <rakesh@sptci.com>
To: "users@jackrabbit.apache.org" <users@jackrabbit.apache.org>
In-Reply-To: <C6D199AB-F24F-42D9-A91E-D7F323E2C13D@maya-systems.com>
Content-Type: text/plain;
	charset=utf-8;
	format=flowed;
	delsp=yes
Content-Transfer-Encoding: quoted-printable
Mime-Version: 1.0 (iPhone Mail 7E18)
Subject: Re: getRootNode() takes 27 seconds...
Date: Fri, 26 Feb 2010 14:40:47 -0600
References: <CD0413B2-1DF1-42DC-95EC-C196C16ACE58@maya-systems.com>
 <3A0E5276-0580-44B9-8D04-8B055455DE58@sptci.com>
 <96DFE183-932B-4673-B3C7-0E9B91C87E48@maya-systems.com>
 <C19CAE78-C0AD-49F1-B1ED-88089D41095A@sptci.com>
 <C5FD086A-FA16-40FD-9F95-52CDE126AFB1@maya-systems.com>
 <C6D199AB-F24F-42D9-A91E-D7F323E2C13D@maya-systems.com>

Is there any difference in performance between session.getRootNode() =20
and session.getNode("/")?

Rajesh

On Feb 26, 2010, at 2:04 PM, Fran=C3=A7ois Cassistat =
<f@maya-systems.com> =20
wrote:

> Hmm, this is weird.
>
> If I use session.getNode("/path/to/some/node"); at start, it works =20
> in 2 seconds
>
> If I use session.getNode("/"); it takes 25 seconds...
>
> Under /, there is only two nodes (jcr:system) and my application =20
> root node.
>
> Once found, I get the root in 0ms, but maybe it is cached.
>
>
> F
>
>
> Le 2010-02-26 =C3=A0 2:26 PM, Fran=C3=A7ois Cassistat a =C3=A9crit :
>
>> Rakesh,
>>
>> Each time I create a connection and get the root node.
>>
>> After that, each access to a node seems a little slow, so querying =20=

>> and fetching the 1500 nodes described below take one or two =20
>> minutes, but maybe decreasing the number of nodes could help. But =20
>> here, nothing as dramatic as 27 seconds for one node.
>>
>>
>> F
>>
>>
>> Le 2010-02-26 =C3=A0 2:14 PM, Rakesh Vidyadharan a =C3=A9crit :
>>
>>>
>>> On 26 Feb 2010, at 12:51, Fran=C3=A7ois Cassistat wrote:
>>>
>>>> Thanks Rakesh for your reply,
>>>>
>>>> actually root have only one node, below 2 or 3 levels of that of =20=

>>>> that there is one node with 1500 nodes, this is the maximum. All =20=

>>>> other nodes have between 0 and 10 nodes.
>>>>
>>>> 1500 may be a lot and I may implements a tree strategy to =20
>>>> decrease the number of child of this node. But does that sub-sub-=20=

>>>> sub-sub-node make the getRootNode() going that slow?
>>>
>>> No, that should not affect the root node retrieval.  Is the time =20
>>> to access repeatable, or just the first time?
>>>
>>> Rakesh
>>>
>>>>
>>>>
>>>> F
>>>>
>>>>
>>>> Le 2010-02-26 =C3=A0 1:16 PM, Rakesh Vidyadharan a =C3=A9crit :
>>>>
>>>>>
>>>>> On 26 Feb 2010, at 11:32, Fran=C3=A7ois Cassistat wrote:
>>>>>
>>>>>> I have inserted ~25K nodes with ~1500 files (about 1.6G of =20
>>>>>> files) and now Jackrabbit takes 27 seconds to get the root =20
>>>>>> node, is it normal? Maybe it is because am I using the =20
>>>>>> standalone server (2.0)? Also I am making tests since jcr2spi.
>>>>>>
>>>>>> I'm a newbie with JackRabbit configuration files, so I did not =20=

>>>>>> make much modifications to repository.xml
>>>>>
>>>>> In my experience the biggest issue is the number of child nodes =20=

>>>>> under a parent (and not repository configuration).  The general =20=

>>>>> recommendation is to try and keep child nodes low (1000 does not =20=

>>>>> seem too bad in my experience), but not 25K under a single parent.
>>>>>
>>>>> Rakesh
>>>>
>>>
>>>
>>
>