La descarga está en progreso. Por favor, espere

La descarga está en progreso. Por favor, espere

                                 .

Presentaciones similares


Presentación del tema: "                                 ."— Transcripción de la presentación:

1                                  

2 Bases de datos de alta disponibilidad
David Motta Sales Consultant Oracle Mexico

3 Reto de Disponibilidad en Servidores
El Servidor es el único punto de falla Servidor 1 Instancia ‘A’ Now that we’ve compared the cost of clusters, let’s look at how high availability differs. In the traditional, single server configuration, if, for any reason, the server goes down (due to a power failure, server crash or hardware failure), the Oracle databaase instance on the server is not available. [Click for slide build] The server acts as a single point of failure for the availability of your Oracle database. Your challenge lies in re-starting your server as quickly as possible. But what do you do if your physical server can’t restart? A pre-requisite of restarting your Oracle database is an active server. If you have hardware problems, that recovery could take a while. [Click for next slide] Base de Datos ‘A’

4 Clusters sin RAC Se evita que el Servidor sea el único punto de falla
Instancia ‘A’ Servidor 2 One solution is to use a clustered server environment to remove the server as a single point of failure. In fact, this is the typical scenario where clusters have been used in the past to provide higher availability of your database applications. Using a cluster, you remove the server as a single point of failure. You have a second server available to take over in the event of a failure to your primary server (or Server1 in this example). But you’ll note that your database only operates on one of the servers in your cluster at any point in time. [Click for next slide] Base de datos Se evita que el Servidor sea el único punto de falla

5 Clusters with ‘Cold’ Failover
Reiniciar la instancia en Server 2 Servidor 1 Servidor 2 Instancia ‘A’ Should Server1 fail, your Oracle9i Database can easily be restarted on Server2. With this architecture, your Oracle9i Database is only available on one of the servers in the cluster at any point in time. This form of cluster addresses the challenge of server failure only - it does NOT provide any additional scalability as you cannot scale beyond the limitations of one server in your cluster. This type of cluster is referred to as a ‘cold’ cluster, since you are effectively using the facilities of the clustered hardware and software to protect from server failure. And although this ‘cold failover’ process protects against the failure of an individual server, the process of restarting the database and failing the users over to a remaining live server requires some time when your database will not be available. [Click for next slide] Base de datos

6 Real Application Clusters
Servidor 1 Instancia ‘A’ Servidor 2 Instancia ‘B’ Al fallar la instancia 1, la Base de datos permanece disponible With Real Application Clusters, every server in the cluster is actively participating as part of the clustered database. Unlike the ‘cold’ cluster we just examined, with Oracle9i Real Applications Clusters, every server in your cluster has full access to your database. [Click to next build] If one of the servers in your cluster fail, your users are automatically and transparently failed over to a remaining live server. And with Oracle9i Real Application Clusters, this failover operation can occur in seconds - effectively masking server failures from your users. Remember with Oracle9i Real Application Clusters, your shared database is already up and running on the other servers in your cluster. Base de datos Se protege de fallas de Servidor

7 Escalar con Clusters

8 Real Application Clusters
Empezar en pequeño, crecer incrementalmente Escalable y altamente disponible Agregar capacidad bajo demanda sin Tiempo fuera de servicio

9 Pagar primero y escalar después
3 6 9 12 15 18 21 24 Meses 300% Carga de Trabajo 200% 100%

10 Escalar con Oracle 3 6 9 12 15 18 21 24 Meses 300% 200% 100%
Carga de Trabajo 200% 100%

11 Scale Out Clusters Escalar aplicaciones bajo demanda
Eliminar Hardware ocioso Facil migración – no se requieren cambios en la aplicacion Incremento en la confiabilidad

12 ¿¿¿Qué sigue??? Presentación nueva plataforma Oracle
Escribir 10gmexico

13                                  


Descargar ppt "                                 ."

Presentaciones similares


Anuncios Google