At some point every IT team ventures out on a mission to consolidate its storage assets. Over the years the organization probably has created a collection of systems that do a variety of tasks ranging from systems for high-performance storage to systems for backup and systems for archive. The challenge with any consolidation mission is it almost always fails. At some point, sooner rather than later, a new workload shows up that just won’t fit into the consolidation strategy.
Storage sprawl, the inverse of consolidation, typically occurs when adding new workloads. Usually, the current storage system lacks enough performance or capacity to meet the new workloads demands. Or even more common, there is a fear that adding the new workload to the existing system will cause other workloads to perform badly.
Can SDS Solve the Storage Consolidation Problem?
One method to not encounter the problems associated with storage consolidation is not to attempt to consolidate at all. The organization could go with a software-defined storage (SDS) solution that provides a standard management interface across disparate systems. Some SDS solutions even offer tiering to lower cost storage system. The problem with this approach is while the software management is centralized, the data center does still need to manage multiple storage hardware silos. Each hardware platform has its own upgrade path and method, as well as interface through support contract and vendors.
Can a Single Box Solve the Storage Consolidation Problem
Another method to solve the storage consolidation problem is to look for a single hardware platform, driven by software that can address the various workloads that may cause contention for data centers. From a hardware perspective, the system would need to support flash drives and high capacity hard disk drives to provide both fast access to active data and affordable storage for long-term data retention. From a software perspective, it would have to, at a minimum, offer both NAS and block storage protocols. It should also support object storage.
To truly consolidate storage, the system would also have to provide data protection and archive. More than likely, an organization won’t backup or archive to the same system as their primary storage system. But if that hardware and software for primary storage, backup, and archive all were from the same vendor with the same software, data management as well as hardware management becomes easier, even across multiple system.
It is important to clarify that for this consolidation to work, the customer realizes that the primary storage, backup storage, and archive storage are the same hardware and storage software. The only component that changes is the drive configuration of the hardware with the primary storage system being all or mostly flash storage and the backup and archive systems being all or mostly hard disk storage.
The storage software would need some additions made for the backup and archive use cases. Snapshot and replication features built into most storage software already can create backups. The copy on the backup system would need a different and deep snapshot retention schedule that the primary storage system.
The more dramatic change to the storage software is the need for archiving functionality. Archives could be sent to the same systems as backups, or it could be an independent system. But the software will need features to migrate data from the primary storage system, and it will need to institute compliance and retention features, which may include a WORM only volume capability as well as a data verification facility.
Consolidation needs to be redefined. Instead of trying to consolidate to a single system, IT planners should look to consolidate down to a small group of systems all driven by the same software with the appropriate features to cover the backup and archive uses cases. The result is simplified hardware management and consistent data management throughout a workload’s life cycle.