Getting more intelligent about storage management.
Until now, however, information intelligence was viewed by the storage community as a closed, storage-only proposition. Storage vendors added such features as virtualization, storage resource management and time stamping to their products in an effort to add "intelligence" and value. While this has certainly improved the tools available, it is too narrow a view.
The customer is now demanding a smarter way to address storage intelligence--a transformation that will come about through viewing storage as a service of the overall needs of the business rather than an island unto itself. This approach allows organizations to simplify, securely manage and protect information and resources in a manner optimally aligned with business objectives. If that sounds to you like ILM-based practices, it is. Storage intelligence is a key service of a good ILM-based management practice.
In this article, we take a look at why intelligence is so important to storage management and answer the age-old riddle--should that intelligence reside on the switch, the application, the server, the array, on a virtualization layer, or somewhere else?
Touching the Business
When storage is assessed in terms of the business as a whole, it becomes clear that its key value is housing and protecting critical information assets. That is not to say that storage is an insignificant part of the organization. After all, storage is the only place in IT that touches 100 percent of the business information. Thus, storage becomes a critical element of any strategy to enhance operational efficiency, bring about compliance, manage risk, or adequately secure business information.
That's why it is no longer feasible to just "throw storage at the problem," (i.e. add more arrays, bigger disks or yet another SAN). The end result of that approach is an unwieldy and cost-prohibitive storage pool that can become difficult to manage, producing an evironment in many storage networks where all information is managed as if it has equal importance to the business.
What is called for is a move away from the brick and mortar storage concept. Building a storage network and dumping all data into one vast storage pool doesn't tell you anything about the value of that information. While it may be substantially easier to manage as a single pool, this is not just a storage problem. Information intelligence is also required. By being more intelligent about information and its value to the business, we can better protect, manage, secure, make available, and move information so it is always at the right place at the right time in a way that delivers real value back to the business.
The Role of ILM
ILM is a management practice that shifts the setting of policies and service requirements up to a collaborative team comprised of the business group, records managers, I.T. managers, and the company's security or risk managers. This is where the 'value of information' to the business becomes the deciding factor. Information becomes the central actor in the way the infrastructure is operated. Storage and data services now play an important role in fulfilling the service requirements of any class of information over its lifecycle. Intelligent storage management plays a strategic role in creating effective ILM practices. How? Consider these four steps as the process for getting started:
1. Identify key business processes and information, data, networking and storage resources (services and assets)
Before you can make the right determinations on how to protect, place, or provision performance or availability for data with your storage systems (RAID, mirroring, replication, etc.), you must first identify the business practices, data information, and security services and data and storage assets you currently have. This typically includes identifying those activities that are critical to your business, the data required and generated by those activities, and the servers and other devices available to store that data. An accurate audit of the current SAN, as well as an in-depth GAP analysis, enables the development of the ideal design for the unique needs of the business. Further, the data gathered proves invaluable in the development of best practices that can be rapidly implemented in the storage environment.
2. Classify the information based on its business value and the classify the capabilities of the storage resources
Once resources and information assets are identified, the next step is to evaluate the business value of the information and secondarily the capabilities of available storage assets and data services. The plan is to match both in setting policies and service level objectives.
3. Define storage policies and practices to support business requirements and implementing the appropriate storage infrastructure
Now you can define policies and the supporting processes that establish best practices for handling your information assets appropriately based on business value. Only by implementing steps 1 and 2 thoroughly is it possible to accurately place information on the most cost-effective storage media. Further, you can then assign information the applicable level of protection, security, and performance based on your assessment of its value to the business.
4. Automating the defined policies and processes
Repetitive manual processes, however, bleed all the efficiency out of even a well-designed system. Therefore, automation of specific tasks becomes a vital element of storage intelligence. By automatically taking action based on established policies, tedious and repetitive tasks are eliminated and costs are significantly lowered.
Not, however, that automation is not an end in itself. It should only be applied where it actually adds value--by eliminating tedious storage "grunt work," by taking care of routine backup tasks, or by buying the storage administrator more time to take a more strategic or proactive role.
Most vendors, by now, realize that greater intelligence is the wave of the future. Where they differ, though, is in the exact location of that intelligence. Switch vendors make a good case on why intelligence should be placed in the switch itself. Array vendors offer an equally well-crafted argument on the wisdom of putting the intelligence in the device controller. Application vendors, too, counter that the application should control all, with the hardware elements being made to align to the needs of the software. And then, of course, virtualization specialists advocate moving the intelligence a step above the array, the switch or the application in order to permit the rapid exchange and management of information in a heterogeneous environment.
So who is right? The answer is all of them. Each part of the storage infrastructure must be intelligent enough to manage its role in the grand scheme of things. Intelligence does and must reside in the switch to manage SAN flows, intelligence should reside in or near the array to enable provisioning in a just-in-time model. SAN management intelligence must reside in an appliance or via software to enable management of information to drive business continuity. All of these components must work in concert to enable intelligent management of information.
Hardware intelligence, while vital, must be subordinated to the bigger picture of the overall management of that information. The concept of intelligent management actually encompasses far more than storage assets, and extends into business process optimization. To deliver true business value, it has to stretch beyond the storage environment to include user management, network management, storage management, access management, and security management. Such cross-disciplinary smarts currently lie largely beyond existing storage products, but this is the direction we need to move--and rapidly.
The good news is that progress is being made in this direction. Some storage management vendors have made great strides in increasing the scope of their management capabilities. The Storage Networking Industry Association (SNIA), too, is working hard on this issue. The SMI-S standard, for example, is a good starting point. It started out as physical storage management and continues to progress with feature sets for data management. For example, SMI-S version 1.1 provides copy services and the future version 1.2 is planned to include ILM specific services.
But standards won't do it alone. What is required is for vendors to work together, collaborating, with each other to produce products that can and do make the grade when it comes to delivering business value. And that means taking storage to the next level, where it is one part, though a pivotal part, of the broader IT landscape.
Eventually, we will arrive at where we need to be. By correlating and aligning the information to broad business needs we will create storage environments that add true value to individual business segments as well as the business as a whole.
Jim Geronaitis is vice president of BrightStor storage management at CA and a member of the board of directors at the Storage Networking Industry Association (SNIA).
Opening shots in continuing stories ...
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||first in/first out|
|Publication:||Computer Technology Review|
|Date:||Aug 1, 2005|
|Previous Article:||Avamar defuses data explosion: company tackles data growth disease, not symptoms.|
|Next Article:||Emulex switches gears to grow beyond HBAs.|