Steps to Improve Storage Management
Many of today’s CIOs and IT professionals in the storage management industry are making a shift from silo-driven IT operations centers that focus on technology to becoming service providers for their businesses and customers. In order to successfully pull off such a shift, they have to synchronize IT and business operations, uncover exactly what existing business needs there are and meet them without decreasing the service levels currently provided.
This shift usually means reducing money spent on IT infrastructure maintenance to free up resources to drive innovation so business units can reap the benefits of additional services and capabilities. “The challenge is on the one hand, live with cost reduction in the management of your infrastructure and, at the same time, add new services and make it even more complex,” said Thomas Goepel, worldwide business strategy and portfolio manager, HP StorageWorks Services.
Goepel said a lot of customers in the storage management space don’t have a single-vendor storage environment. They have multiple-vendor storage environments that are typically set up as islands. “A lot of customers look at storage as a separate entity from a management perspective than the rest of the IT infrastructure, and when you look at all of the components it makes it very, very inefficient. I think the key challenge that our customers are facing right now is, is there a way of transforming that individual silo approach to management into a consistent, overarching IT management that not only looks at the storage piece, but is tied into the management of the rest of the environment: the server, the networking, the application management? You want to create a storage management environment that’s heterogeneous and can give you a single complete view.”
Examining or evaluating storage only when you’re trying to solve business issues regarding availability or performance can make it tough to identify and locate problems when they crop up in the management space. “Storage management is not just about making sure your device works,” Goepel explained. “From a management perspective, it’s all about how can you monitor performance if there are certain actions happening. Can you set thresholds that trigger a certain action automatically rather than manually? And then how can you be sure that, when you’re running an application, that transaction is going form that application to the storage and back fulfilled in a certain amount of time?
“You also need to prove that you’re improving storage capabilities, etc, which means you need reporting capabilities that give you a clue of how storage management is really doing. Are you meeting your service levels? On top of that, you want to manage availability and fast recovery so when something happens you can recover from the incident and return to a normal-operations state quickly,” Goepel said. “You want to look into incident management, back-up management, recovery management, those kinds of things.”
In order to improve your organization’s storage management platform take proactive actions. Don’t wait until there’s a problem and then try to find out where the mistake is. Look at the system not simply as storage management but as unified infrastructure management. “You need to have the tie into other areas because storage cannot be treated as a stand-alone piece,” Goepel said. “The end user only cares that his application is up and running; he doesn’t really care if it’s a server problem or a storage problem. You as the IT admin need to look at where is the problem coming from, and if you can’t establish that link between the server, the application, the network and the storage, you’re going to be searching forever for the problem.”
In order to create a more efficient storage management system, start with the basics. Create a single environment to manage your multi-vendor storage system. This will improve the efficiency of the storage management professional running the environment. “Consider the device level management or things like provisioning of storage resources,” Goepel said. “This is incident management on the individual device level. The next step is taking this into a reporting function. Can you assign the storage resources and create reports? You’ll have to change the way you’re managing from incident management to providing service levels to a business owner or department in your organization. The third step is to take that link into the application world. How do I ensure that my Oracle database that runs on my disc array is really performing the way it should? Are my beta sets laid out on my disc arrays in an optimum way, yes or no? How can I ensure that the storage held on my disc array is performing the way it should? That’s taking the critical applications and creating inside modules that give you the availability and performance of those environments so you can react faster to challenges that come from the application world into your storage world.”