Showing results for 
Search instead for 
Did you mean: 

Veritas InfoScale Enterprise for Oracle Single Instance with Fast Failover

Level 2

Today, many businesses use Oracle Real Application Clusters (RAC) to build scalable Oracle database clusters. This feature comes at a considerable added cost. For many companies with high availability and uptime requirements, they can mitigate the cost by deploying Oracle Single Instances that uses InfoScale Enterprise with Fast Failover. I want to highlight how InfoScale Enterprise can help protect your Oracle landscape and accelerate database performance.   

Unlike standard clustering solutions, InfoScale Enterprise with Fast Failover facilitates the rapid failover of Oracle databases by providing a shared storage system that dramatically reduces failover time. Additionally, this solution provides automation, orchestration, high availability management, and fault detection while significantly reducing costs compared to native Oracle functionality. 

Figure 1 Veritas InfoScale Enterprise for Oracle Single Instance with Fast Failover OverviewFigure 1 Veritas InfoScale Enterprise for Oracle Single Instance with Fast Failover Overview

There are two core concepts to Oracle Single Instance with Fast Failover. The first is that the Oracle database is stored on an InfoScale Enterprise cluster where the data is available and accessible by all nodes at the same time. The second is that Veritas Cluster Server (VCS) is responsible for monitoring, orchestrating, and automating the failover process between the active and passive nodes. See Figure 1. 

The fast failover of a single instance Oracle Database minimizes failover time by eliminating the extremely time-consuming operations within the failover process such as unmount, remount, and other processes required to connect and access the filesystems. By avoiding these operations, fast failover can also reduce maintenance downtime.   

The fast failover of a single instance Oracle Database can take several seconds or more depending on various parameters, including the size of the database and other system configuration details. 

Veritas Cluster Server enables high availability by providing near-instant fault detection at the kernel level and is responsible for initiating failover operations. Combined with customizable Oracle DB deep level monitoring, VCS agents act immediately to take corrective actions in the event of a system fault.  

Shared Storage 

Veritas InfoScale Enterprise contains 2 features that can help build a shared storage infrastructure (in the absence of a SAN): Flexible Storage Sharing (FSS) and Cluster File System (CFS). FSS enables local storage devices contained on each node of a cluster to be shared with the other cluster nodes, and CFS enables you to mount a file system on multiple nodes at the same time.  


Figure 2 Shared Storage Created with Flexible Storage Sharing and Cluster File SystemFigure 2 Shared Storage Created with Flexible Storage Sharing and Cluster File System

FSS creates the shared storage system, making use of direct-attached storage or SAN devices connected locally to the server. That is, the locally accessed disk devices are shared through FSS with the other nodes in the cluster, creating a pool of disk devices. Logical volumes can be created from the shared disks, on which the CFS is created. All nodes then mount the CFS in the cluster at the same time.  

FSS supports disk devices in the public cloud, making it possible to build a shared storage infrastructure in Amazon AWS, Microsoft Azure, and Google Cloud Platform. 


InfoScale Enterprise also allows for significant performance benefits.  

  • Oracle Disk Manager (ODM) extensions are supported by InfoScale Enterprise, which enables Online Transaction Processing performance equal to raw disk partitions, but with the manageability benefits of a file system.  
  • The SmartIO feature enables granular and intelligent caching at the application level by caching read operations inside of the server. SmartIO can also provide caching services via Flexible Storage Sharing by extending locally cached data to other nodes within the cluster. SmartIO is also an ideal tool for accelerating database read performance in Public Cloud deployments without having to pay extra for higher performance disk services. 
  • Dynamic Multipathing (DMP) maximizes storage performance by load-balancing I/O activity across all available paths in all major hardware RAID array products. DMP also reduces the total cost of ownership by eliminating the need for third-party multipathing software.   

Protecting Your Data 

InfoScale Enterprise features Storage Checkpoint technology that is disk and I/O efficient in creating persistent images or snapshots of the filesystem. Storage Checkpoints significantly reduce the amount of storage used, compared to full filesystem copies of data required to create backups or clones. In addition, Storage Checkpoints enable you to restore on-disk backups with rollbacks for quick restoration of data.  

Data can also be moved offsite to another InfoScale and Oracle cluster for redundancy and high availability with Veritas Volume Replication (VVR). VVR enables cost-effective replication of data over IP networks. VVR provides organizations with extremely flexible storage hardware independent alternatives to traditional array-based replication solutions as well as a robust mechanism for moving data into public cloud environments. 

InfoScale Enterprise makes Oracle databases highly available and resilient to disruption on-premises and in public clouds with native support for Amazon AWS, Microsoft Azure, and Google Cloud Platform. Using VVR and the VCS Global Cluster Option, you can design your InfoScale and Oracle infrastructure to be highly available between datacenters and public clouds.


Figure 3 Multicloud ExampleFigure 3 Multicloud Example

For more information on Veritas InfoScale Enterprise with Oracle Single Instance databases with Fast Failover, please see the following whitepapers:  

For general technical information on Veritas InfoScale Enterprise, please read the following: