Arnold Here's what Kurt (the jUDDI project lead) had to say about this.
"The war should be kept container agnostic. We are pretty clear what to do to configure a database:http://juddi.apache.org/docs/3.x/userguide/html/chap-database_setup.html Also we have the tomcat-distro to get going. We should point users to both these things when this question comes up." On Fri, Jan 3, 2014 at 7:02 AM, Alex O'Ree <spyhunte...@gmail.com> wrote: > Here's a few pointers > juddiv3.war should be deployed exploded, meaning that it should be > unzipped. The file structure should look like > tomcat/webapps/juddiv3.war/WEB-INF > tomcat/webapps/juddiv3.war/META-INF > tomcat/webapps/juddiv3.war/(more stuff goes here > > Please send in the log files if possible or at least any stack traces > to help narrow down why it didn't just work > > For your other questions > 1) hibernate, in general, is a bit faster than openjpa and is sort of > a more commercialized implementation of jpa. Many people prefer it, > but juddi works with both of them. > 2) There should be a context.xml in juddi3v.war/META-INF. I just > checked the distribution and it looks like it wasn't in the bin/war > file. However it is in the distribution zip. Grab is from here: > juddi-distro-3.1.5\juddi-tomcat-3.1.5\webapps\juddiv3\META-INF. Note: > this is probably the reason it failed on start up and it is a bug > we'll have to address > 3) The database driver probably should just go in juddiv3.war/WEB-INF/lib > 4) If the username/password provided for OpenJPA/Hibernate can create > a database, I believe it will be done automatically. If you wanted to > create it yourself, the default name is juddiDatabase > > On Fri, Jan 3, 2014 at 4:00 AM, Arnold Harbers > <arnold.harb...@crv4all.com> wrote: >> Thanks Alex for your response. >> I am trying to understand this. >> >> I have downloaded juddi-distro-3.1.5.zip. >> From this distribution I have extracted juddiv3.war from the bin directory >> and deployed it. >> This resulted in my initial error. The guide for the database connection >> setup states that "By default jUDDI uses an embedded Derby database. This >> allows us to build a downloadable distribution that works out-of-the-box, >> without having to do any database setup work." Does this mean that I did >> something wrong in the deployment? Should I be able to get juddi up and >> running (using Derby) in an existing development environment that already >> uses Tomcat? If so, I think that should be my first step and after that I >> should try to switch database connection. >> >> Following the guide for the database connection setup: >> 1. edit persistence.xml. In the juddiv3.war it is the OpenJPA version. In >> the oracle section the hibernate version is mentioned. Should I replace the >> one in the juddiv3.war and use the Hibernate version? >> 2a. define datasource in context.xml. There is no context.xml in the >> juddiv3.war. Do I need to copy it myself to the directory >> <tomcat>/juddiv3/META-INF/ directory? >> 2b. There are some context.xml examples for MySql, Postgress, but not for >> Oracle. Do you have an example for Oracle? >> 3. add the database specific driver to your classpath. What is the classpath >> in a Tomcat environment? Where do I have to put the file? >> 4. I assume I have to create the database juddiv3 as well? >> >> With your answers I will try to get a successful implementation with an >> Oracle backend and update the oracle installation guide. >> >> Thanks, >> Arnold >> This message is subject to the following E-mail Disclaimer. >> (http://www.crv4all.com/disclaimer-email/) CRV Holding B.V. seats according >> to the articles of association in Arnhem, Dutch trade number 09125050.
