To support your business applications and business commitments, you need a stable and robust backend infrastructure. Backend infrastructure consists of common components like network, storage, backup and disaster recovery (DR) solutions, along with management and monitoring tools.
For continued storage data availability, storage vendors can provide capabilities to protect your business in the event of a drive failure by introducing a separate, stand-alone drive in the hardware solution that’s referred to as a “hot spare drive.” This drive is on standby for use in the event of a functional drive failure. In such a case, it automatically replaces the failed drive and hence assists with business uptime requirements.
For IBM SAN Volume Controller, a solution built with IBM Spectrum Virtualize software, IBM extends data protection further by not only providing a hot spare drive solution but also introducing a “hot spare node” (a SAN Volume Controller node).
How the IBM solution works
In traditional scenarios, if one SAN Volume Controller node is not operational (whether for known or unknown reasons), applications can still continue with input/output (I/O) using the I/O group partner node, but with degraded performance. This scenario, however, presents challenges for the business until corrective action is taken (for example, the replacement of a faulty component or invoking your DR).
But the hot spare node can address this challenge. If a SAN Volume Controller node fails or is down due to scheduled maintenance, the hot spare node stands in for the failed or serviced node and supports the performance requirement as expected from two active-active controllers. The key enabling technology behind the hot spare node capability in the IBM Spectrum Virtualize code for SAN Volume Controller is NPIV (N-port ID virtualization). NPIV provides a virtual worldwide port name (WWPN), which can move freely from one SAN Volume Controller node (the failing one) to first the I/O group partner, and then the hot spare node, and back. Furthermore, it does this with minimal disruption to the host accessing volume(s) from that node.
(For a more technical deep dive, please see “IBM Spectrum Virtualize N_port ID Virtualization for improved availability.”)
Figure 1: Hot spare node online role during node failure situation
IBM Systems Lab Services has implemented such a solution for a number of clients, including a major bank in the Middle East and a government entity in South Africa. In both cases, our solution addressed high availability needs for Spectrum Virtualize with resiliency tests in a controlled and unrestrained manner. And in both situations, a hot spare node functioned exactly as expected.
Are you interested in such a solution?
For clients running SAN Volume Controller, IBM Systems Lab Services can provide a solution for introducing hot spare nodes for better high availability for your business without compromising on performance requirements while a failed controller is recovered or replaced.
Lab Services offers infrastructure services to help you build the foundation of a smart enterprise. Our Storage and Software Defined Infrastructure consultants perform onsite infrastructure services and offer deep technical expertise, valuable tools and successful methodologies to help you solve your business challenges, gain new skills and apply best practices. Contact us today.