High availability requires the following hardware and system components:
Two homogenous, networked systems, preferably with similar configurations in terms of resources like CPU, memory, and so on. Adaptive Server supports VCS on RedHat Enterprise Linux 2.1 Advanced Server and VCS version 2.1.
You should also install the VCS GUI (graphical user interface) to facilitate configuration and administration.
The two systems must have access to shared multihost disks, which store the databases for the Adaptive Server configured for high availability.
Install Veritas Volume Manager 3.2 to manage disks and create resources like DiskGroup and Volume.
Use third-party vendor mirroring for media failure protection.
Create a service group on each system. A service group is a set of resources that provides a specific service. To provide a service for an Adaptive Server that is configured for high availability, the service group should include such resources as DiskGroup, Volume, Mount, IP, NIC, and so on for Adaptive Server. A sample service group and the resource dependency graph is shown in Figure 10-1. See the Veritas Cluster Server User's Guide for more information on how to create a service group and how to add resources to a service group.
Each service group must contain at least two resources
with one resource of type HAase for VCS 2.1.
Use the cluster command to establish resource dependency so that
the resource type depends on the other resources.
Configure both public and private networks on both the nodes.
See your hardware and operating system documentation for information about installing platform-specific high availability software.
In Figure 10-1, the configuration of the service group has one DiskGroup, syb_vrtsdg1, on which four volumes are created. One volume is for the Adaptive Server installation, one is for databases that are created on the file system, and the other two are for databases created on raw devices. The two mount resources are created for file system of type vxfs (Veritas file system) layering on the volume resources. The resource, syb_ase125 of type HAase is the Adaptive Server installation, which sits on top of the mount resources. syb_ase125 also requires resource IP, which also requires resource NIC for public network access.
The service group SybASE runs on the primary node and another service group, SybASE2 (not included in Figure 10-1) runs on the secondary node, with a similar configuration.
Figure 10-1: Sample service group running on Veritas Cluster Server
Copyright © 2004. Sybase Inc. All rights reserved. |
![]() |