|< Day Day Up >|| |
So far, we've offered some exposure to internal database features that aid you in the struggle for availability. You have been through the setup and administration of a RAC environment. You have confirmed your credentials as a true paranoid by configuring a Data Guard disaster protection scheme. You've got your media backups humming along with RMAN, and you've tested out the Flashback technologies to save you from your users. All of these solutions armor you against downtime, working together to provide a complete high-availability solution for your database.
But there's this other thing. This kinda new thing. This kinda cool thing. It does not act like any of the HA technologies we've espoused to this point. It can complement any of them, or be used in their stead. It provides a slightly different approach to availability that emphasizes distribution of database activities. Oracle calls it Streams.
Streams, as a term, actually applies to a larger set of functionality within Oracle that has nothing to do, at its core, with availability. Streams is really a queuing technology, referred to frequently as advanced queuing. The queuing nature of Streams resembles any queuing architecture that you may already have been exposed to: it is a publish/subscribe approach to event management. But on top of the queue structure, you can create a robust and extremely advanced solution for data sharing and load balancing that distributes data across multiple independent and available databases.
When used in this fashion, Streams typically gets called Streams replication, and while it is primarily a data-sharing technology, it can be used as a highly customizable availability solution. But we should make something very clear: Streams is not an 'out of the box' experience. Streams requires a larger ongoing commitment to the specifics of your environment than any other HA solution that we have described this far. However, with a little elbow grease and some ingenuity, Streams may just be the future of your HA world.
Given the complexity of Streams, we will keep things very simple in this chapter, and cover the primary topics of importance where Streams can be used for availability. For all other Streams topics, we defer to Oracle's complete Streams documentation set as referenced in the bibliography.
|< Day Day Up >|| |