Storage Insights

How Data Centers Can Benefit From SSD Today

Mon, 7 Mar 2011 11:13:23 -0600

Almost all storage managers in medium to large sized data centers are keeping their eyes on solid state storage (SSD). While discussion about the technology is ubiquitous, actual implementation has been limited to a finite set of niche use cases, usually high end databases. For increased adoption to occur solid state storage has to be delivered in such a way that it can be implemented more broadly across a wider range of applications, so that its benefits can be experienced by more of the enterprise.

What is a Virtualization Performance Specialist?

Fri, 4 Mar 2011 05:54:38 -0600

Server virtualization is certainly widespread, but while the initial results have been promising, most IT personnel feel that the benefits of virtualization could be increased. For many, they’re not seeing the performance, consolidation, and resource utilization levels they had expected. The reasons for this vary but in general effectively optimizing storage and SAN I/O resources to fully exploit a virtual server infrastructure is much harder than many realize. Server administrators simply don’t know enough about storage issues and typically don’t have a view into storage I/O related metrics. And, the skill sets of traditional storage administrators may not support what is essentially a new paradigm. What may be needed is a new focus or responsibility for existing server and storage IT personnel, call it the Virtualization Performance Specialist, that deals with the nuance and complexity of this increasingly interrelated environment.

Using Primary Storage Deduplication to Address the Data Affordability Gap

Tue, 1 Mar 2011 12:03:46 -0600

Data is growing, it’s a fact of life in IT organizations. Budgets are growing too, and significant percentages of current and projected IT spending are being earmarked for more storage capacity. However, there are some details that should be understood. Year over year data growth is currently more than 50% (and increasing) while IT budget growth is less than 10% (usually much less). Even with continuous reductions in price per GB, the costs of new capacity required is growing a lot faster than the availability of funding. In response to this, storage manufacturers are using primary deduplication to address this Data Affordability Gap.

While data growth is well accepted in the storage industry, what may be surprising is that the rate of that growth is also growing. In physics, a change in velocity is called acceleration; you could say that data is also accelerating. Some of the drivers of this are the rise in social networking (at work), mobile computing, compliance, the increased use of images in everything (with more resolution), data mining and analytics. Again, the point here isn’t just to establish that data is growing, but that data growth itself is growing.

Achieving Application Aware Availability

Mon, 28 Feb 2011 09:12:01 -0600

In a recent article, Storage Switzerland discussed methods for protecting an application from a storage system failure. Storage hardware, along with server hardware, is increasing in its reliability. The growing culprit of application downtime is the application itself either through application lock-up, data corruption or poor performance. The problem is that most availability solutions focus on the hardware, the operating system and the network connections - not the application. Any attempt to achieve true application availability should now include an intelligence and understanding about the application itself. In short, true availability should include application awareness.

Intelligent Flash Storage Demands Native Data Services

Wed, 23 Feb 2011 10:58:10 -0600

In the most recent article in our “Intelligent Flash Storage Demands an Innovative File System and Native Data Services” series, we covered the importance of a flash-optimized file system. It lays the foundation for the data services like thin provisioning and snapshots. Many of these data services are considered critical capabilities that allow storage managers to do their jobs. As with the file system, these data services should be optimized to exploit a flash-only sustainable storage system.

Intelligent Flash Storage Demands an Innovative File System

Thu, 17 Feb 2011 07:32:02 -0600

The move to a flash-only storage system will bring many benefits in the areas of performance, energy efficiency and data center floor space savings, but the gains cannot come at the expense of capabilities that the enterprise now considers necessities. As discussed in part one of our series “Intelligent Flash Storage Demands an Innovative File System and Native Data Services” not only should a flash-only system support these necessities they should also allow them to exploit the full capabilities of an ultra-low latency flash environment. A key aspect of this and the foundation for the other software services is the file system.

Proficient Object-based Replication keeps up with High Performance Storage Systems

Wed, 16 Feb 2011 11:23:25 -0600

Replication is a data service that has become a standard feature in most storage systems, often used to move data off-site as part of a disaster recovery strategy. But demands for increasing storage growth are starting to outpace the abilities of traditional replication technologies and the WAN connectivity they use to transmit data. Similar to the benefits that object-based architectures brought to file systems, object-based replication has the proficiency to keep up with these high growth and high performance storage systems.

Intelligent Flash Storage Demands an Innovative File System and Native Data Services

Tue, 15 Feb 2011 07:59:24 -0600

As the cost of flash memory continues to decrease, enterprises are looking to deploy them in more horizontal applications. Many of today’s solid state storage systems are niche solutions only focused on solving unique performance bottlenecks in specialized applications, such as databases, OLTP, and data warehousing. The fact that these solid state storage systems lack of data services, like snapshots, deduplication, NAS features, and replication, narrows their deployment use case potential. Attempting to integrate flash arrays with existing enterprise storage arrays that provide those data management and protection features is also challenging.

How Should Primary Deduplication Be Delivered?

Mon, 14 Feb 2011 09:22:45 -0600

Primary storage deduplication is becoming a necessary feature for storage managers. How primary storage deduplication should be delivered is a question that’s now on the top of many vendor product managers’ agendas. Their choices for delivering primary storage deduplication vary from proprietary software implemented directly into the architecture, to an external appliance, to a third party API set like that available from Permabit Technology. The choices vendors make directly impact end-user deployments, the extent of their product’s ROI and its long term flexibility.

The promise of reduced capacity requirements, its associated costs and the ability to ‘bend’ the storage growth curve as well as the increased efficiency of other operations like snapshots, replication and backup copies all make primary storage deduplication compelling. While there is no longer debate on the value of the technology, vendors are struggling with which of the above methods to use to integrate deduplication into their primary storage solutions.

Scaleable NFS Acceleration with Enterprise Solid State Cache

Thu, 10 Feb 2011 08:01:50 -0600

As application servers that perform build, render or search functions grow they often develop front end request issues and back end storage I/O performance issues. The front end request problem is most often created by many users accessing the application at the same time and is handled by developing a scale out application server architecture. The back end performance problem of dealing with the storage I/O needs of those build, render and search requests is only compounded by adding dozens of additional application servers, which allow even more users to make those request.

The Value Of Highly Available Backups

Wed, 9 Feb 2011 08:49:15 -0600

The backup process is often considered a secondary process. Because of the number of things that can go wrong, missing a backup or having the completion of the process delayed by a few hours is something the IT Managers have learned to live with. The problem with this is that when something does go wrong, especially due to a backup device failure, it causes a ripple effect that’s felt throughout the data center. Steps to build redundancy in the backup process have been expensive but ineffective. Backup virtualization should be looked at as an alternative to standard backup techniques to gain the value of highly available backups.

Arming The Cloud With Storage

Tue, 8 Feb 2011 07:27:51 -0600

Storage needs for cloud providers has typically been broken into two groups; storage for Cloud Compute and storage for Cloud Storage. Cloud compute data centers often need scalable, high-performance SAN-based storage since most of their actual processing is done locally within the data center. Cloud storage data centers typically need scalable but inexpensive storage, since most of the processing is done outside of the data center and the storage in the data center is just a repository. Many cloud providers actually have both needs and have to use two separate storage systems to meet user demands. What providers are looking for though is a single system that can satisfy both usage scenarios.

Using Clones To Manage VMware Storage Growth

Fri, 4 Feb 2011 08:53:33 -0600

VMware server virtualization through its abilities to consolidate servers, increases operational efficiencies and improves an organization’s ability to prepare for a disaster, bringing an unprecedented level of return on investment. However, these gains can be quickly eaten away by uncontrolled storage infrastructure expenses. While other concerns include performance and data protection, as discussed in the article “IOPs are More Important Than Air”, front and center for most organizations is curtailing capacity growth.

Top Criteria For MSP Storage

Thu, 3 Feb 2011 07:41:59 -0600

Managed Service Providers (MSP) have a unique business challenge. Instead of dealing with the problems of a single data center, they have to deal with the problems of potentially hundreds or even thousands of virtual data centers with ever changing workloads and priorities all consolidated under one roof. The storage system that an MSP selects can be either a key enabler for the business or it can significantly hold the business back, in many cases meaning the difference between profitability and bankruptcy. Storage systems, like those from Pillar Data Systems, are almost tailor made to meet the top criteria for MSP storage and allow storage to be an asset to the organization instead of a liability.

Enterprise SSD Requires RAS (Reliability, Availability And Serviceability)

Wed, 2 Feb 2011 08:19:13 -0600

Today Solid State Disk (SSD) is viewed by most storage managers as a niche solution to solve specific storage performance problems. For SSD to move out of the niche category and become a more mainstream storage consideration for most of the active data in the enterprise, it must address the reliability, availability and serviceability (RAS) concerns that those managers have about the technology. Solid-State-only storage systems, also called “Sustainable Storage”, like those from Nimbus Data Systems, are addressing this requirement.