Got it. If you only need connections, you can write your own provider class which returns connections (or DataSources) based on a name-to-database mapping of your chosing.
Or, if you really must provide each connection as a JNDI DataSource, I would investigate writing your own JNDI provider that operates on the single, manual connection, and returns DataSources based on the information found via that connection. You can borrow some code from javaranch.com's jrunittesthelper, a package which provides JNDI lookups for JUnit test cases. Tim -----Original Message----- From: Rob Gregory [mailto:[EMAIL PROTECTED] Sent: Monday, January 23, 2006 4:28 AM To: Tomcat Users List Subject: RE: Setting up connection pools "on the fly"... Hi Tim, Thanks for the assistance. Our basic requirement is during tomcat load/start-up to query the database (with a single manual connection) and read the available database environments that have been pre-configured. Then create these dynamically so they are available as if they had been configured within conf/../context.xml etc. I managed to get this kind of working buy rewriting the config details during start-up but it seems at this time it's too late for Tomcat to use these (i.e. the connection pools are already set up). They would become available the next time tomcat was restarted... I then tried adding the config files as a 'watched' resource but this lead to some strange behaviour. While in most cases we don't change servers as such - we do provide our own tools to manage the database details such as username, password etc and these need to updated without modifying the config files directly. Hopes this clarifies the problem a little (shout if you need further or alternative explanations). Thanks again for your help. Rob -----Original Message----- From: Tim Lucia [mailto:[EMAIL PROTECTED] Sent: 22 January 2006 23:42 To: 'Tomcat Users List' Subject: RE: Setting up connection pools "on the fly"... How often do you change servers? What you describe below can be handled by editing the appropriate context / resource and restarting Tomcat. Perhaps more detail on the business requirement, rather then stating "I must define them on the fly" would enable us to give you a better solution. -----Original Message----- From: Rob Gregory [mailto:[EMAIL PROTECTED] Sent: Sunday, January 22, 2006 5:41 PM To: 'Tomcat Users List' Subject: RE: Setting up connection pools "on the fly"... "One doesn't typically add new database servers "on the fly", and this option would require defining a new JNDI datasource for a new server" That's exactly what I need to be able to do... Is this possible on Tomcat start-up as it doesn't have to strictly on the fly just on start-up. I mainly need to remove the configuration details from the context.xml file and read these from the database before tomcat starts and creates the pools. Any help much appreciated. Rob -----Original Message----- From: Tim Lucia [mailto:[EMAIL PROTECTED] Sent: 22 January 2006 14:21 To: 'Tomcat Users List' Subject: RE: Setting up connection pools "on the fly"... The point of connection pooling is to eliminate the overhead of setting up and tearing down a (TCP, database, AAA) connection for every database transaction (typically, the web request in a web app.) This can add 100s or 1000s of milliseconds to every request, and is quite expensive. If you can architect your application so that there is a JNDI data source for each server (database server, instance, etc.) with its own pool of connections, you can always call setCatalog (http://java.sun.com/j2se/1.5.0/docs/api/java/sql/Connection.html#setCat alog (java.lang.String)) on the connection returned from the pool. One doesn't typically add new database servers "on the fly", and this option would require defining a new JNDI datasource for a new server. Of course all catalogs processed in this way must be accesible to the user specified in the connnection pool properties. I would suggest setting the defaultCatalog property as well, so that you are not accidentally connected to the last-used catalog from the previous borrower of the connection. Tim -----Original Message----- From: Rob Gregory [mailto:[EMAIL PROTECTED] Sent: Sunday, January 22, 2006 8:23 AM To: 'Tomcat Users List' Subject: RE: Setting up connection pools "on the fly"... I also have the same requirement (to provide connection pools on the fly) but also need these pools available to Cocoon (so need them to map to the standard jdbc/mypool jndi location. Tomcat declares it's initial context as being readonly so I have no idea how to go about binding the pools I have created dynamically. Any pointers would be a great help. Thanks in advance Rob -----Original Message----- From: Alex Turner [mailto:[EMAIL PROTECTED] Sent: 20 January 2006 22:50 To: Tomcat Users List Subject: Re: Setting up connection pools "on the fly"... Maybe you don't even want a connection pool, rather a simple non-pooling datasource. If you generate a pool of connections for each and every database, you may end up with a great many open connections to your database server that hardly ever get used. I don't know too much about MySQL, but in most other databases, each connection has certain allocated resources, and the database typically limits the max connections. I have a similar situation, but I use a simple datasource, and a servlet to create a new datasource for each database on initialization based on a database entries. You can easily create a datasource on the fly in a servlet. I don't know if this is the 'right' answer, but it works for me. Alex On 1/20/06, Warrick Wilson <[EMAIL PROTECTED]> wrote: > I've got a project where there's a basic interface, with some basic > database access. However, when someone "signs up", then a new MySQL > database will be allocated for them. There's security associated with > it, and the user will have to log in to identify himself. > > How can I create a new connection pool for that new database for my > web app? My current pool is set up in context.xml - do I modify that > on the fly from my web app? > > What I'm doing currently is figuring out who the user is and which > database he's associated with, and when I do any database work, the > first statement is "USE <databasename>" to get the proper database. > > Is there a better approach? More like a proper approach? (Says he who > isn't sure he fully understands connection pools yet....) > > -- > Warrick Wilson > [EMAIL PROTECTED] > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
