Designing Storage for High Availability
Storage high availability depends on redundant hardware, alternate access paths, and software failover working together to keep applications connected during faults and maintenance.
Designing storage for high availability means preserving application access when a component or connection becomes unavailable. Redundant controllers are one part of that design: hosts need working alternate paths, and software services must also remain available. Seagate Exos capabilities should therefore be evaluated within the complete access path, against specific failure and maintenance requirements.
Summary
- Array redundancy supports availability at the storage layer. Host connections and software services must also support continued application access.
- Multiple connections need supported failover behavior. A second cable alone does not establish how a host responds to a failed path.
- Failover preserves access through an alternate component or path. Rebuilds restore drive protection; neither replaces backup.
- Before deployment, check how storage maintains access during controller, connection, and software-service failures, as well as component maintenance.
What Does High Availability Require From Enterprise Storage?
A useful availability requirement names the service users need and the events it must withstand. “Keep the storage online” leaves too much unspecified: an array can remain powered while a host loses its connection or a file service becomes inaccessible.
Start with the application’s access requirement. Identify the data it uses, the systems that serve it, and the interruption the business can tolerate. Then define the faults and planned maintenance activities the design must accommodate. A failed drive, unavailable controller, and disconnected host path affect different parts of the system.
These requirements guide enterprise storage architecture more effectively than a hardware feature list alone.
Storage Availability Depends on Controllers, Paths, and Software
Array Controllers and Serviceable Components
Exos PROTECT capabilities include active-active dual controllers, hardware RAID with ADAPT, and redundant, hot-swappable components. These features give the array ways to protect data and support continued service when individual components require attention.
Identify which components are redundant, how the system responds when one becomes unavailable, and what work is required to restore normal operation. Hot-swappable hardware supports replacement while the system operates, but it does not establish that every maintenance procedure is interruption-free.
The choice between SCALE and PROTECT also affects responsibility. SCALE provides capacity infrastructure for a selected storage stack; PROTECT supplies array-level RAID services. Availability planning must follow the architecture actually being deployed.
Host Connections and Alternate Paths
Redundant storage controllers are useful only if the host can continue reaching the data. Trace the connections between each host and the array, including the adapters and any intervening infrastructure. Determine what happens when one connection fails and how an alternate path becomes active.
The historical Seagate and ATTO MultiPath Director solution brief illustrates this responsibility. It describes automatic failover and failback across Fibre Channel paths, alongside tools for managing and monitoring those paths. Those are host-connectivity functions, distinct from RAID inside the array.
Use that example to identify what needs evaluation, not as proof of compatibility with a current deployment. Confirm the selected adapters, operating system, drivers, and storage configuration together. Multiple physical connections do not establish that the required multipathing behavior is configured or supported.
File-System and Application Services
Storage access can depend on software services above the array. If users reach data through a file-system controller or file server, that service needs its own availability design. An operational array cannot compensate for every failure elsewhere in the access chain.
Seagate’s historical Ardis solution brief makes the distinction concrete. The Exos X array used active-active storage controllers, while the Ardis Dual HA AVFSHead used active-passive file-system controllers. A standby AVFS controller could take over the file-system role. These were complementary layers in a documented combined solution, not interchangeable descriptions of the same redundancy feature.
The example supports a planning question: which service must take over when the active one fails? It does not establish current Ardis compatibility or guarantee transparent failover for another software stack.
Failover Preserves Access; Rebuilds Restore Storage Protection
Failover and rebuilds address different problems. Failover preserves access through an alternate component or path. A rebuild restores protection after a drive failure. A system may remain accessible while protection is being restored, so both the service state and the recovery process matter.
Evaluate ADAPT rebuild planning separately from controller and path failover. Neither redundancy nor a successful rebuild removes the need for backup and recovery arrangements.
Seagate’s Kyoraku Sangyo case study describes a historical Exos X/E configuration with redundant RAID controllers and power supplies. The account also says continued operation during multiple drive failures depended on conditions. That qualification matters: a customer configuration illustrates a design, but cannot establish a universal failure tolerance or uptime guarantee for current Exos systems.
Storage Availability Checklist
Before putting storage into service, check how the proposed configuration handles failures and maintenance.
- Confirm applications retain access when a controller becomes unavailable.
- Check that an alternate connection works when the primary path fails.
- Identify which software service takes over if the active service fails.
- Confirm how component replacement affects access and data protection.
- Document how the team detects faults and restores normal operation.
Evaluate the complete configuration, not just the array’s advertised capabilities.
Plan Exos Storage Availability With ASI
Storage high availability requires array protection, host paths, and software services to work together. Exos redundancy supplies part of that design; the complete configuration must meet the application’s access requirements during the faults and maintenance events the business needs to tolerate.
Bring ASI the host-path diagram, proposed storage software, and failure scenarios the customer needs to tolerate. Its account managers and pre-sales technical team can help frame the Exos hardware discussion around those requirements. Assign responsibility for validating failover across the complete configuration before committing to application availability.
Frequently Asked Questions
Do Dual Controllers Guarantee Application Availability?
No. Dual controllers provide redundancy at the array layer. Hosts still need supported access paths, and the software services used by applications must remain available. Assess the complete configuration against defined failure scenarios.
Is Multipathing the Same as RAID?
No. Multipathing manages alternate connections between hosts and storage. RAID protects data within the storage configuration. A design may need both because path failures and drive failures affect different responsibilities.
What Does a Five-Nines-Class Design Claim Establish?
Seagate’s five-nines-class description characterizes the PROTECT design. It is not, by itself, a measured uptime result or contractual guarantee for an application using that array. Evaluate the full configuration and any applicable service commitments separately.
Find the Right Storage for Your Next Project
Talk with an ASI storage expert about Seagate Exos options that fit your workload, capacity, and growth requirements. As an Authorized Distributor, ASI helps you narrow hardware choices, coordinate component sourcing, and identify configuration questions for Seagate’s product experts. Share your needs to start planning your next storage purchase.