PRIMECLUSTER Installation and Administration Guide 4.1 (for Solaris(TM) Operating System)
|
Contents
Index
![Previous](../pclsecom/previous.gif)
|
2.3.2 Scalable Operation
This section explains the topologies used for scalable operation:
Scalable
Definition
- A cluster system consists of two or more operating nodes, and all the nodes are used for online cluster applications. This operation mode is suitable for parallel jobs that use the I/O load balancing and load sharing on a parallel database.
Advantages
- Throughput can be improved for a cluster application by expanding nodes.
- Degenerated operation is possible even if part of the cluster applications stops.
Disadvantage
- If part of the cluster applications stops, throughput of the cluster applications cannot be guaranteed because degenerated operation is assumed.
Failover image
![](../pclsesrc/p010101i.gif)
![](../pclsesrc/boxchuui.gif)
Scalable operation can be used in combination with some PRIMECLUSTER-related products. For information on the related products, see the "PRIMECLUSTER Product List. "
High-availability scalable operation
Definition
- Refers to the topology in which standby operation is configured for each cluster application that constitutes scalable operation. Suitable for a parallel database for which scalability and availability are required, as well as parallel job execution for which load share/load balance is used.
- Standby operation that constitutes scalable operation can be combined with 1:1 standby and N:1 standby, with priority transfer.
Advantages
- Even if failover occurs in one of the cluster applications that constitute scalable operation, the throughput of all the cluster applications can be maintained by using a redundant configuration.
- Throughput of the cluster applications can be improved with node expansion.
- Degenerated operation is possible even if part of the cluster applications stops.
Disadvantage
- Since a redundant configuration is used, the node usage efficiency is poor during normal operation.
Failover image
The following illustrates failover when two 1:1 standby operations are combined to enable scalable operation.
![](../pclsesrc/p010101j.gif)
![](../pclsesrc/boxchuui.gif)
High-availability scalable operation can be used in combination with some PRIMECLUSTER-related products. For information on the related products, see the "PRIMECLUSTER Product List. "
All Rights Reserved, Copyright (C) FUJITSU LIMITED 2005