IJC Deployment Guide

This document provides a guide to deploying Instant JChem to your users. Most aspects are described in more detail in other parts of this administrators guide. This document provides a basic overview of the options available.

Computer requirements

IJC runs on the desktop computers used by your users. Windows, Linux and Mac OSX operating systems are supported. Other operating systems providing a Sun Java 1.6 JRE (or higher) are also probably possible.

IJC provides an in-built local database (Derby ) that can be used for personal use. This provides unrestricted functionality within the local database. The local database cannot be used to share data amongst multiple users.

If sharing and collaboration of data amongst multiple users is required then you need to provide a database server for your users to access. This server runs only the database software (Oracle and MySQL are supported) and does not run the IJC software. This database server is not provided with IJC and must be installed separately. Free versions of both databases are available. This server should be a relatively fast machine with sufficient memory and you are strongly advised to have a fast network connection between the database server and the client computers on which IJC is running as IJC is a very data intensive application.

Whether you are running a local database or a remote (shared) database much of the computationally intensive operations (such as structure searching) are performed on the IJC client, so these machines should be fast computers with sufficient memory, especially when using large data sets. For handling up to a million structures we would recommend a relatively recent computer (e.g. one bought in the last 3 years) and allocate at least 200MB RAM to IJC. For larger datasets you need correspondingly more memory, and for very large datasets you should consider using the JChem cartridge for Oracle as this moves the memory requirements of structure searching from the client machine to the JChem server that is being used by the Oracle cartridge.

For notes about each particular database type look here .

Deployment options

IJC comes in a form with a standard installer. This allows you to install on each of your computers. However if you need to manage a large number of users this becomes impractical. The advantage of the installer version is that it needs no supporting infrastructure and you can add extension modules easily to it, and so it is easier to customise.

For managing a large number of user the Java Web Start version is more suitable. This allows each user to launch IJC by clicking on a link in a web page and the IJC software is loaded and run using Java Web Start. To be able to use the Java Web Start version all you need is a standard web server on which you place the software and the launch page. This is described fully here .

It is also possible to deploy fully configured projects to all your users. This avoids the need for them to set up database connections, and allows collaboration and sharing of data in a shared database. This is most useful when combined with using Java Web Start deployment, but is also possible with the installer based versions. This is described fully here .

Once you have a shared project you can refer to data in it from outside and inside IJC using URLs. This, for instance lets you send links by email that open IJC and show the results of a particular query. To do this you need a servlet compliant web server such as Tomcat or Jetty (or an equivalent running as part of a full-blown app server such as JBoss, Glassfish or WebSphere). This is described fully here .

Problems?

If you have any questions or any of this is unclear then please report this to the IJC forum .