By Shreedhar Ganapathy on Mar 31, 2007
We are looking for feedback from users of NetBeans and GlassFish on what they would like to see in terms of their runtime interactions through the IDE with cluster(s) in GlassFish.
Most IDEs provide the ability to work with a single instance of an application server such as GlassFish, etc. I do not see the notion of cluster(s) in the IDEs unless I have missed something obvious. There must be various use cases wherein a developer (typically an advanced developer/deployer/tester such as a System Integrator) would need the ability to interact with a cluster through the IDE.
Clustering typically involved an environment consisting of many instances that can be located on different remote machines. Such instances are created/started/stopped/deleted on each machine by a dedicated agent process that follows instructions from an administrative server within which a cluster is a logical configuration entity while instances are physical entities. So, within the IDE, there are possibly but a few operations that can be executed/managed.
So here are some questions for which I'd appreciate responses through comments (Add more if you can think of any further):
- Is there indeed a need for interacting with clusters through the IDE?
- What problems do you see today with respect to clusters from within your IDE?
- If the browser based admin console of the appserver can be lauched from within the IDE would this be sufficient to address your needs?
- If not, why is the IDE an important piece to help solve that problem?
- If interacting with cluster(s) is important, what should that interaction functionality be from within an IDE?
- If debugging seems like an obvious use case, would that not already be addressed by working with individual instances that are part of a cluster?
- What other use cases do you work with that you wish an IDE could provide when working with clusters?
Do add more questions and answers if you can think of more.