Paul, sorry my misunderstanding - just comments at the moment as I work through the code.
Norman On Jan 4, 2008 10:24 AM, Paul Fremantle <[EMAIL PROTECTED]> wrote: > Norman > > I only meant that the DB would store the BLOB in a separate file, not > that it would automatically do diffs. > > I agree that we need to implement the diffs. > > Paul > > > Norman Barker wrote: > > Paul, > > > > I think if you are running a cluster then the sys admin will have > > already considered this and perhaps implemented GFS so a filesystem is > > available. > > > > Looking at the version table > > > > CREATE TABLE VERSIONS ( > > AID INTEGER NOT NULL, > > VN INTEGER NOT NULL, > > CONTENT BLOB, > > AUTHOR VARCHAR (500), > > UPDATED_TIME TIMESTAMP, > > UNIQUE (AID,VN), > > FOREIGN KEY (AID) REFERENCES ARTIFACTS (AID)); > > > > even if the DB implements diff I am not sure how this schemas takes > > advantage of this. > > > > Not being deliberately argumentative here, but the whole geospatial > > use case is based on imagery (large) and this just isn't stored in the > > DB. > > > > > > Norman > > > > On Jan 4, 2008 10:00 AM, Paul Fremantle <[EMAIL PROTECTED]> wrote: > >> Norman > >> > >> We discussed the use of a separate store for the content. However, many > >> databases (DB2, Oracle) implement this already. Also, there is then a > >> significant problem in deployment. For example, if you have a cluster, > >> you now have to have a clustered file system as well as database. > >> > >> I like the idea of doing diffs as a pluggable idea. > >> > >> Paul > >> > >> > >> Norman Barker wrote: > >>> Hi, > >>> > >>> again correct me if I am wrong since I am new to the code, but looking > >>> at the database schema I see that the complete content of the resource > >>> is being stored in the artifact table, and in the version table. > >>> > >>> Really I would prefer the content to be stored outside of the DB (and > >>> this what I am putting into the ORM version) as use case for data is > >>> large (potentially gigabytes per resource). For the version table is > >>> it on the road to support xdelta type diffs (as in Subversion). I see > >>> there is a library called JRCS which was in Apache Commons and is now > >>> LGPL. > >>> > >>> I mention this since it isn't just applicable to the ORM version, and > >>> does impact the scalability of the registry. > >>> > >>> Norman > >>> > >>> _______________________________________________ > >>> Registry-dev mailing list > >>> [email protected] > >>> http://wso2.org/cgi-bin/mailman/listinfo/registry-dev > >>> > >> -- > >> Paul Fremantle > >> Co-Founder and VP of Technical Sales, WSO2 > >> OASIS WS-RX TC Co-chair > >> > >> Office: +1 646 290 8050 > >> Cell: +44 798 447 4618 > >> > >> blog: http://pzf.fremantle.org > >> [EMAIL PROTECTED] > >> > >> "Oxygenating the Web Service Platform", www.wso2.com > >> > >> _______________________________________________ > >> Registry-dev mailing list > >> [email protected] > >> http://wso2.org/cgi-bin/mailman/listinfo/registry-dev > >> > > > > _______________________________________________ > > Registry-dev mailing list > > [email protected] > > http://wso2.org/cgi-bin/mailman/listinfo/registry-dev > > > > -- > Paul Fremantle > Co-Founder and VP of Technical Sales, WSO2 > OASIS WS-RX TC Co-chair > > Office: +1 646 290 8050 > Cell: +44 798 447 4618 > > blog: http://pzf.fremantle.org > [EMAIL PROTECTED] > > "Oxygenating the Web Service Platform", www.wso2.com > > _______________________________________________ > Registry-dev mailing list > [email protected] > http://wso2.org/cgi-bin/mailman/listinfo/registry-dev > _______________________________________________ Registry-dev mailing list [email protected] http://wso2.org/cgi-bin/mailman/listinfo/registry-dev
