A New Storage Architecture that Eliminates the Need to Ever Make a Forklift Upgrade Again
Ask any business owner or IT administrator how much storage they will need in a few years and they will likely hem and haw trying to come up with a reasonable answer. Ask them to share their true feelings and they will in all likelihood respond, "I don't know." The good news is that storage architectures are now available that take the risk out of this uncertainty and that do not require yet another expensive, disruptive and risky forklift upgrade.
Corporate data is being created at unprecedented levels, and the pace is increasing in unpredictable ways. According to one recent multi-university study of 26 corporations and large non-profits volume of data is expanding by 35%-50% every year in most organizations which essentially doubles the volume of data stored every two years. The highest rate observed was 150%. The researchers made the sobering observation that in every organization studied, data growth far exceeded revenue growth.1
The Bigger Storage Challenge
Anticipating how much storage any one business will need to store all of this data is a significant challenge. However the bigger challenge is that legacy storage architectures simply were not designed to support the loads associated with the dramatic increases in data that organizations will experience over the next several years. As such, organizations need more than new storage products; they need new storage architectures.
This thought may well result in groans from business and IT professionals alike who remember the pain and disruption associated with their most recent forklift storage system upgrade. But implementing a new storage architecture does not have to involve expensive business disruptions and painful storage migrations.
New Storage Architecture minus the Storage Pain
To offer the advantages of scale-out storage without the traditional headaches, new software-defined storage architectures have emerged that address today's storage pain points while simultaneously creating a solid storage foundation for tomorrow.
Specifically, software-defined grid architectures are now available that provide the advantages of a highly scalable storage infrastructure without the compromises in performance, capacity and cost associated with legacy storage architectures. These grid architectures enable any size organization to create a robust storage infrastructure that is simple, powerful, efficient and affordable.
The Grid is a Powerful Storage Architecture
A software-defined grid storage architecture scales all performance elements proportionally as it is architected for fault-tolerance rather than failover. Each new storage node also adds CPU, RAM and network bandwidth to the storage pool.
Gridstore, a leading provider of these next generation software-defined storage architecture, leverages client CPU resources to encode the data before it is sent over the network. By doing the encoding first, storage nodes need to do very little work other than simply storing and retrieving data, maximizing resource efficiencies. This permits high performance storage nodes to be built from affordable, energy-efficient components.
In Gridstore's Grid solution, its vController software seamlessly installs on any network device that uses the Grid. This eliminates the storage controller as a bottleneck, and gives every device direct access to all storage nodes. The result is a highly parallel, high performance file IO environment that gets more powerful with the addition of each storage node--without the expense of over-provisioned storage controllers or 10Gb Ethernet networks.
Gridstore also uses a data protection technology that is superior to RAID called erasure coding. Erasure coding provides exponentially better fault-tolerance than RAID, enabling file operations to continue uninterrupted and with no data loss even in the event of multiple simultaneous storage node failures. Erasure coding is also significantly more space efficient than Replicas commonly used in scale-out storage architectures (300% overhead for Replicas vs 30% overhead for Erasure Encoding to offer the same level of protection).
Easy, Affordable, Non-Disruptive Installation and Management
Gridstore is designed for non-disruptive entry into an environment and for non-disruptive scaling through automated recognition of new storage nodes as they are added to the Grid. This is important because any kind of disruption costs money.
Initial installation of Gridstore storage is non-disruptive to IT operation for the following reasons:
- Its vController software may be installed without rebooting the server.
- It may be deployed without requiring specialized storage skills
- It completes quickly so large blocks of IT staff time are not needed
- Integrates seamlessly with existing AD security groups and backup software
Ongoing management of Gridstore storage is also non-disruptive.
- It automates configuration, provisioning and healing so no IT staff heroics are required to maintain the Grid
- Painful forklift upgrades are eliminated as the Grid may be scaled by simply plugging in another storage node
- New storage nodes are recognized and added to the Grid storage pool automatically.
A software-defined storage architecture is affordable because it consists of inexpensive building blocks. There is no need to purchase a system designed for storage needs three years down the road which forces organizations to plan and budget for the unknown. Instead organizations only pay for what they need now, and when their needs grow they may add storage incrementally and inexpensively at the time it is needed.
Because of the low incremental costs, such purchases fit into standard corporate purchasing processes rather than requiring special review and approval from the CFO. This frees up both IT staff and corporate management time while simultaneously enabling the storage environment to be more responsive to changing business needs.
Backup and Recovery Ideal Starting Point
Backup and recovery is an ideal starting point for Grid deployments. The Grid allows organizations to build a scalable, efficient and highly reliable backup solution. Also, it seamlessly integrates with leading backup software vendors such as Veeam and Symantec.
Whether an organization is migrating away from tape, or simply looking for a more efficient and affordable storage solution, Gridstore works well as a backup target since it offers:
- Higher performance for backup and restore operations.
- More fault-tolerance so backups complete even in the event of hardware failures
- Higher availability so backup jobs always complete within the backup windows.
- Power efficiency with each storage node only consuming 25-35 watts.
- Scalability. As the size of backups grows so does the speed and capacity of the Grid; keeping the backup/restore environment unified, simple, efficient and reliable.
In using Gridstore as his backup target, Eddie Palmer, a system administrator with Micro Matic, says, "Gridstore was literally 'set it and forget it.' Backups just work and there is no worry about tapes. It works completely with our existing infrastructure."
Gridstore storage provides performance suitable for all types of file IO at a price that makes it a cost-effective replacement for tape or other more expensive backup targets. Adding Gridstore to a data center may also save money by extending the life of existing primary storage--offloading the IO, bandwidth and storage capacity associated with file data to the Grid.
A Perfect Fit for Today's Dynamic Data Growth
The storage architectures of yesteryear are a poor fit for today's dynamic data growth. The unpredictable growth rate of data in most businesses means that they can no longer accurately forecast their storage requirements or reasonably budget for expensive system upgrades.
This dynamic environment dictates the need for a new storage solution architected to emerge. This solution must be properly architected so it scales affordably, incrementally and non-disruptively.
The Gridstore Grid fulfills these next-generation storage architecture requirements. It turns simple building blocks into powerful storage, deploys quickly and non-disruptively without requiring specialized storage skills, and scales performance with no architectural bottlenecks. Perhaps most importantly, it is architected to meet today's storage requirements while also providing businesses with a solution that can grow both storage capacity and performance to meet future business needs--without breaking the budget.
1 Beath, Cynthia, Irma Becerra-Fernandez, Jeanne Ross, and James Short. "Finding Value in the Information Explosion." MIT Sloan Management Review RSS. N.p., 19 June 2012. Web. 31 Dec. 2012.