From users-return-10652-apmail-jackrabbit-users-archive=jackrabbit.apache.org@jackrabbit.apache.org Tue Feb 17 08:19:45 2009 Return-Path: Delivered-To: apmail-jackrabbit-users-archive@minotaur.apache.org Received: (qmail 61856 invoked from network); 17 Feb 2009 08:19:45 -0000 Received: from hermes.apache.org (HELO mail.apache.org) (140.211.11.2) by minotaur.apache.org with SMTP; 17 Feb 2009 08:19:45 -0000 Received: (qmail 23032 invoked by uid 500); 17 Feb 2009 08:19:43 -0000 Delivered-To: apmail-jackrabbit-users-archive@jackrabbit.apache.org Received: (qmail 23016 invoked by uid 500); 17 Feb 2009 08:19:43 -0000 Mailing-List: contact users-help@jackrabbit.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: users@jackrabbit.apache.org Delivered-To: mailing list users@jackrabbit.apache.org Received: (qmail 23005 invoked by uid 99); 17 Feb 2009 08:19:43 -0000 Received: from athena.apache.org (HELO athena.apache.org) (140.211.11.136) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 17 Feb 2009 00:19:43 -0800 X-ASF-Spam-Status: No, hits=2.2 required=10.0 tests=HTML_MESSAGE,SPF_PASS X-Spam-Check-By: apache.org Received-SPF: pass (athena.apache.org: domain of imadhusudhanan@zohocorp.com designates 72.5.230.103 as permitted sender) Received: from [72.5.230.103] (HELO svwall.zoho.com) (72.5.230.103) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 17 Feb 2009 08:19:34 +0000 Received: from 172.31.250.201 (172.31.250.201 [172.31.250.201]) by svwall.zoho.com with SMTP id 1234858752555631.7865068625326; Tue, 17 Feb 2009 00:19:12 -0800 (PST) Message-ID: <11f83513929.5698538339416868605.-8577824408014023880@zohocorp.com> Date: Tue, 17 Feb 2009 13:49:12 +0530 From: imadhusudhanan To: aklimets@day.com Subject: Re: Re: Moving to DFS System .. Cc: users@jackrabbit.apache.org In-Reply-To: Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_19940_231204463.1234858752296" References: <11f5eb84519.-6944610101884563522.3289008303588878124@zohocorp.com> <510143ac0902100124n7104149ex2fda57cf7bd07452@mail.gmail.com> <11f606e630d.-6433629101906078105.-828846402884277074@zohocorp.com> <11f7f86fa21.-6823704681498889486.101557615757859726@zohocorp.com> X-Priority: 3 (Medium) User-Agent: Zoho Mail X-Mailer: Zoho Mail Status: RO X-Virus-Checked: Checked by ClamAV on apache.org ------=_Part_19940_231204463.1234858752296 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Hi Alex, There are two options that might work for you, but both involve some=20 coding effort: one is to use Jackrabbit's WebDAV server "library" to=20 build your own server-side implementation of WebDAV that connects to a=20 Hadoop FileSystem.=20 I under stand this as I got to define the classes under org.apache.jack= rabbit.wedbav.xxx packages which will now be compatible to HadoopFileSystem= . Correct ?? The other option would be to implement the full JCR=20 API via Jackrabbit SPI [2], which is a simpler API than the full JCR=20 API, and build this on top of a Hadoop FileSystem - but this is a=20 rather huge effort.=20 Y u called this to be a huge effort ?? After all i can see very few ap= is in org.apache.jackrabbit.spi package compared to the former to implement= . Do you mean that implementing the existing webdav server library with Had= oop File System will be easier when compared to the spi provision .. ??=20 =20 Also, as far as the scalability is concerned, which one do I prefer= .. ??=20 =20 Regards, MadhuSudhanan I. www.zoho.com 'If you wanna walk quick Walk Alone, if you wanna walk far Walk Together ..= ." ---- On Mon, 16 Feb 2009 Alexander Klimetschek wrote ---= -=20 > Hi,=20 > =20 > (I only answer questions publicly, so this goes back to the Jackrabbit l= ist)=20 > =20 > On Mon, Feb 16, 2009 at 3:39 PM, imadhusudhanan=20 > wrote:=20 > > Currently we use Hadoop API to access files from Distributed File= =20 > > System. I would like to enable Webdav to the same DFS that I use usin= g JR.=20 > > May I know how I can make it possible ??=20 > =20 > Jackrabbit is not a generic WebDAV to file system mapper as you might=20 > think. Since it is a JCR repository, it must allow for all the=20 > fine-grained JCR features (nodes and residual properties, versioning,=20 > node types, locking etc.) that cannot be mapped onto simple OS file=20 > systems or simple filesystem abstractions (as what Hadoop contains for= =20 > example). Theoretically that might be possible, but it's not an option= =20 > for a performant implementation. Therefore Jackabbit has its own=20 > persistence abstraction (mainly around the PersistenceManager=20 > interface [1]), which is driven by the internal architecture to=20 > support the full JCR API.=20 > =20 > [1] http://jackrabbit.apache.org/api/1.5/org/apache/jackrabbit/core/pers= istence/PersistenceManager.html=20 > =20 > =20 > > Also Hadoop DFS has its own FileSystem. I guess that an entry in= =20 > > repository.xml tag will change the file system to what ev= er I=20 > > specify say the org.apache.hadoop.fs.LocalFileSystem etc.=20 > =20 > No, you cannot use it. FileSystem is just a common name for=20 > persistence abstractions, but in this case, Hadoop's FileSystem (base=20 > class org.apache.hadoop.fs.FileSystem) and Jackrabbit's FileSystem=20 > (interface org.apache.jackrabbit.core.fs.FileSystem) are two=20 > completely different things.=20 > =20 > Also, Jackrabbit's FileSystem is somewhat deprecated and today not=20 > used for actual persistence - that's handled by PersistenceManagers=20 > which are at a low-level where they no longer "know" about the=20 > hierarchy, but solely work with uuids and node bundles.=20 > =20 > This means writing a PersistenceManager that works with a Hadoop=20 > FileSystem is probably very difficult or even impossible. Not sure how= =20 > Marcel's implementation works, but it seems to use a different Hadoop=20 > API (not the Filesystem).=20 > =20 > There are two options that might work for you, but both involve some=20 > coding effort: one is to use Jackrabbit's WebDAV server "library" to=20 > build your own server-side implementation of WebDAV that connects to a= =20 > Hadoop FileSystem. The other option would be to implement the full JCR= =20 > API via Jackrabbit SPI [2], which is a simpler API than the full JCR=20 > API, and build this on top of a Hadoop FileSystem - but this is a=20 > rather huge effort.=20 > =20 > http://jackrabbit.apache.org/jackrabbit-spi.html=20 > =20 > Have a look at the following links if you are interested in more=20 > informations about Jackrabbit's architecture:=20 > =20 > http://jackrabbit.apache.org/jackrabbit-architecture.html=20 > http://jackrabbit.apache.org/how-jackrabbit-works.html=20 > http://jackrabbit.apache.org/jackrabbit-configuration.html=20 > =20 > Regards,=20 > Alex=20 > =20 > -- =20 > Alexander Klimetschek=20 > alexander.klimetschek@day.com ------=_Part_19940_231204463.1234858752296--