Optimizing Storage with Global File Virtualization

By Jack Norris, Vice President of Marketing, EMC

Existing storage management methods and tools cannot keep pace with today’s exploding storage requirements.

pplication expansion, proliferation of rich digital media, and stringent compliance requirements all contribute to the fast growing storage demand. According to IDC, storage administrator productivity must increase by 60% per year just to keep up with the anticipated growth in storage capacity.

Traditionally, storage administrators deal with these challenges by procuring excess disk during technology refreshes. While this approach provides a short-term fix, it is inefficient and expensive to maintain as storage grows exponentially. Organizations can simplify NAS management, balance infrastructure utilization, and lower TCO by consolidating their NAS infrastructure using global file virtualization.

The challenge

Organizations are under increasing pressure to dramatically simplify network storage and drive improvements in capacity, performance, and tiered storage management. A potential solution, however, must not introduce any additional risk into storage environments. Administrators do not want to risk data integrity problems, disaster recovery issues or performance bottlenecks.

A related concern for any viable solution is how end-user access is managed. The disadvantages of a solution that requires changing end-user mount points or installing special software on each server or client can significantly outweigh the benefits of uninterrupted access during data movement.

Organizations also need to understand how a solution impacts other management tasks. How difficult is the initial deployment? How does it perform data retention activities? Will a potential solution create management headaches elsewhere or in the future? For example, does a solution provide a proprietary approach that is in conflict with upcoming industry standards?

Benefits of global file virtualization

Global file virtualization not only helps administrators meet the storage management challenges of today, it drives tremendous benefits. A robust file virtualization solution can enable unstructured data management without disrupting end-user or application access. With the right system in place, an organization should experience optimized networked attached storage (NAS), reduced storage management overhead, improved end-user service levels, and additional storage management functionality. The solution should fit seamlessly into heterogeneous storage environments and should optimize networked storage across different vendor platforms, increasing flexibility and lowering TCO.
Storage administrators can look to file virtualization to allow them to easily identify NAS capacity and performance issues. They should be able to identify free resources and initiate file movement without scheduling downtime or disrupting users. Additionally, with the right system in place storage administrators should expect to be able to proactively manage their storage utilization by implementing file archiving or performing a bulk transfer of unused files to a lower tier of storage.
Ultimately, global file virtualization enables organizations to optimize their storage without risking data integrity, requiring downtime, or creating performance bottlenecks as storage capacity continues to grow.

Components of a comprehensive solution

Various monitoring tools are available today that can provide a great deal of information about the status of a network storage environment. A global file virtualization solution provides this information and enables administrators to take action. Following are seven essential functions an organization should expect from a complete file virtualization solution and highlights of EMC’s Rainfinity Global File Virtualization as a specific example:

File management: Policy-based identification and automatic archiving of static files lowers TCO and streamlines back-up and recovery. Rainfinity enables storage administrators to create flexible rules and archiving policies based on attributes such as location, age, size, or type.

Capacity management: A global file virtualization solution should automatically identify over-allocations at the file server, volume or quota tree level and take corrective action. Rainfinity visually presents the most critical capacity issues so administrators can easily identify problems; a simple click allows administrators to analyze problem areas in more detail and resolve issues on demand.

Performance management: Identifying and eliminating performance bottlenecks can be time-consuming and can interrupt user access while data is being moved. Performance management gives administrators visibility into the storage environment and allows them to achieve maximum performance without interrupting data access. Rainfinity identifies file server CPU, volume, and directory process bottlenecks and resolves issues by dynamically distributing content to alternative locations to better balance performance.

Tiered storage management: A tiered storage environment means data resides on the most appropriate platform depending on its value to the organization. A global file virtualization solution should not only identify which data is most critical, but be able to move less critical data to a different tier without interrupting user access. Rainfinity analyzes access and relocates active content, including open files, to the most appropriate storage tier, transparently supporting service levels without disrupting storage systems.

Global namespace management: In a complex, heterogeneous file server environment, it’s critical to centralize the management of Windows and UNIX namespace services to have a truly virtualized environment. Rainfinity's global namespace feature manages and seamlessly integrates with standard namespaces to enable multiple file systems to appear as a single virtual file system.

Migration and consolidation: The ability to move files transparently, without affecting end users and applications, allows organizations to better leverage existing storage and improve ROI. Rainfinity pools multiple file servers while handling the associated complexity surrounding security, permissions, and domains and automating security ID translations and access control settings to ensure that the data, security, and meta data are moved correctly, quickly, and transparently.

Synchronous mirroring: Mirroring data to a remote site provides for rapid recovery so that business can continue in the event of a disaster. Mirroring NAS devices with remote replication augments a local backup strategy to better protect unstructured data. Rainfinity can replicate live production data in multi-vendor environments across IP networks.

Architecture matters

A global file virtualization solution that consists of separate networking, file system, transaction and application processing layers will provide safe, scalable, continuous access of networked storage.

One important consideration is Global Namespace, which provides an abstraction layer that enables end-users to employ a logical name that is separate from the actual physical location. Rainfinity’s architecture rests on the ability to deliver file virtualization without implementing a proprietary namespace that needs to be constantly in the data path. This is accomplished by leveraging industry standard namespaces as a transport mechanism to automatically update client mappings.

With the right architecture, global file virtualization can be deployed easily, scale to support the enterprise, and preserve data integrity.

Non-disruptive deployment

Organizations should expect a global file virtualization solution to seamlessly plug into existing networks without requiring configuration changes to NAS servers, filer servers, clients, application servers, or storage management tools and utilities. It should support all devices that are accessed via industry standard NFS or CIFS protocols, and enable access to both protocols from a single management interface. Additionally, no proprietary client, server or filer hardware or software should be required. One advantage of Rainfinity’s approach is that it is not constantly in-band and does not require proprietary agents to be deployed on every client machine. Rainfinity installation consists of simply plugging the appliance into the network.

Enterprise scalability

Ideally, the global file virtualization solution should be able to move in and out of the data path to provide high-speed throughput and transparent redirection and should remain out-of-band when no transactions are executing. An architecture with an independent network processing layer does not need to reside in the data path, impeding data throughput and response times. Rainfinity’s unique architecture includes advanced multiplexing and de-multiplexing capabilities and network fastpath processing to efficiently process selective traffic.

Enterprise data integrity

Data safety is the most important factor for storage administrators. A scalable and heterogeneous solution is important, but only if the data is well-protected. An independent transaction processing layer manages data mobility so that any system failure during the middle of data movement does not affect data integrity. Rainfinity’s unique transaction-based processing technology provides complete data integrity with no persistent meta data, no single point of failure, and no disaster recovery exposure. Rainfinity manages open files, open locks, and guarantees data integrity at all times.

Summary

Global file virtualization is changing network storage management. With the right solution in place, organizations can easily simplify management, increase utilization, decrease over-provisioning, resolve performance bottlenecks, leverage tiered storage and lower TCO. With the right solution in place, global file virtualization will not only help administrators meet the challenges of today, but can serve as a key building block for utility computing, storage grid strategies and ILM.