High availability is the mantra of today’s data centers. Businesses storing up their digital information are anxious that the data remains highly available and accessible all times thus Business continuity and continuous application availability are among the top requirements for many organizations. They want to reduce latency and disaster recovery within their stipulated “Recovery Time Objective” (RTO) and “Recovery Point Objective” (RPO). This is the laying foundations for setting up “Active-Active” data centers that address these basic needs of the business.

IBM Spectrum Virtualize with its Enhanced Stretched Cluster Configuration (ESC) can supply a rock-solid storage infrastructure for business continuity functions and IBM Spectrum Scale with file system unlimited scalability.

Customer configuration :

The goal of this configuration is to create a solution where :

1. IBM Spectrum Virtualize ESC can handle split brain condition in case of single site failure
2. IBM Spectrum Scale can provide non-disruptive file system access with unlimited scalability through multi protocol access (POSIX,NFS,SMB,OBJECT)
3. Applications continue to run without any disruption while you do maintenance of your storage infrastructure.

Figure 1 : Shows example implementation high-level design
Figure 1 : Shows example implementation high-level design

Customer scenario:

Suppose you have an Application A which is running on spectrum scale using GPFS POSIX/NFS/SMB/Object client interface on Site A and second Application B which is running on Site B.

In this scenario Application A can be served primary by file system whose NSDs are on Site A while NSDs on Site A can have secondary NSDs ( failover NSDs ) node on Site B and same would be the case for Application B which is running on site B. In this configuration if Site A fails , Application A can continue to be served by NSDs on Site B without any disruption because of the IBM Spectrum Scale NSD failover mechanism and vice versa.

This way, customer can run their Application with zero RPO and zero RTO.

For configuration, Please follow the steps mentioned in below redpaper:

More Referenced Links:

For more information about Enhanced Stretched Cluster prerequisites, see IBM SAN Volume Controller Enhanced Stretched Cluster:

For more information about IBM Spectrum Scale quorum and GPFS reliability:

Join The Discussion

Your email address will not be published. Required fields are marked *