Article

Q&A: Where storage management needs to go

Mark Lewis
The promise of administrative bliss and management flexibility are still pie-in-the-sky dreams, says James Geis, director of storage solutions at Forsythe, until standards find their way into management tools. Geis will be addressing storage management in an upcoming conference session at the Spring

Requires Free Membership to View

Storage Decisions conference in Chicago, May 25-27. His session will be focused on the relationship between storage, information, policy and compliance, and how a programmatic approach to policy is the best start to building an efficient information management framework. In this Q&A, Geis gives us a sneak peek at his session and his thoughts on storage management.

What are the key differences between storage management and storage operations? Are they really that different?
Yes, they are different. In fact, I believe that they can be separated into three different facets: discovery, operations and monitoring/reporting.

Discovery allows users to interrogate your entire storage infrastructure and have a central repository of all the elements within the enterprise: individual disks, RAID groups, configurations, controllers, capacities, HBAs, frames, arrays, switches, ports and zones down to the most granular element and the relationships between all the elements. Users want a complete inventory of all the physical and logical elements that comprise the storage infrastructure. It should have the ability to be viewed from any perspective, angle or relationship and allow you to view the information graphically and textually. This last portion would assist organizations in asset tracking/management, maintenance programs, etc.

Operations allows users to manage or control every element within the storage infrastructure; to configure, reconfigure and assign storage within the frames/arrays/devices/volumes, assign/mask LUNs, create zones, manage switches, ports, ISLs, etc. The operational portion would have the ability to interact with servers and host volume management applications. It also would have the ability to use a single GUI that interacts with all APIs, SNMP MIBs, CLIs or other standard interfaces, as well as any application to make/break relationships or control all elements through automation or structured tasks based upon policies. Ideally, the operational facet would allow you to "test" changes prior to actually pulling the trigger and identify "dead" or unused configurations.

Monitoring/reporting allows users to view all statistical data on performance, utilization/allocation, capacities, usage, assignments, benchmarks, metrics, etc. This repository should have the ability to view the data from all different facets and also allow element management/reporting to be integrated with an enterprise management framework for alerts or errors and automated remediation to avoid downtime, unplanned outages or unavailability. Additionally, the ability to perform predictive modeling for all operations to avoid problems would be ideal.

What is the status of storage management standards today? Are vendors reacting with compliant products, and are users buying in?
Storage management standards are certainly gaining traction and all vendors are jumping on the bandwagon. It is required that the storage manufacturers to comply and participate, but the open systems storage area could take a lesson from what mainframe storage management tools have been doing efficiently for decades.

I don't believe the end users will buy in until they see that the standards have been integrated and that the tools decrease the administrative burden and increase administrative options. Each vendor will most likely continue to put their own spin on the standards and we will see another few rounds of the "pass the API" swap. Having standards will greatly enhance the chance of having tools that allow you to manage heterogeneous platforms more efficiently. End users are tired of using disparate tools that do not live up to marketecture (aka vaporware or PowerPoint storage).

After a few years of posturing, virtualization seems to finally be making an impact on storage. Where in the network is virtualization being used most (subsystem, switch/network or host/appliance)? In your opinion, where is right place for it to finally reside?
Virtualization will most likely land in the switch/network. As iSCSI, FCIP and other networked based storage protocols converge onto the TCP/IP network (enterprise applications and databases can now run on NAS for example), we'll be moving away from channel and direct attached protocols. Voice, data and many applications are pushing network infrastructures to be faster and more robust, and as a result, many organizations are reassessing their intranets/internets to provide flexible options for delivery of all information.

Also, as organizations become more decentralized and global in nature, having flexible options for data delivery over the company network is integral for operational excellence. Supporting data over the IP network also provides more options for heterogeneous replication, backup, disaster recovery and business continuity.

While virtualization may be ready for prime time, a lot of users are still building extensive homegrown storage management tools. When can we expect storage management suites from vendors to live up to promise?
Once standards get set, adopted and have a few revisions, and once we really move towards a more TCP/IP-centric data delivery model, we will see some congruity. But we still have a few generational iterations to suffer through. The complicating factor here is that the words storage and information are sometimes used interchangeably. Storage management tools will need to become smarter and not only manage the physical/logical elements of all types of storage (as well as manage replication, backup, restore, aid in recovery, etc.), but they will need to become "information" aware.

We're moving towards a tiered storage model for information delivery (transactional data on higher end storage and reference data on lower end storage, archive and compliance), and until the tool can manage both, we will continue to be frustrated by the available tool suites. Additionally, middleware applications may have to take on the role of understanding storage pools in order to aid in this virtualization process, rather than rely on the storage array/frame, operating system or network. They must be able to see the storage "classes" and match data, performance, location and type with delivery expectations.

What else can attendees of your Storage Decisions 2005 session expect to learn about?
The relationship between storage, information, policy and compliance is becoming more painful every time we read the news. Having a programmatic approach to policy is the best start to building an efficient information management framework.

For more information or an application to the FREE Storage Decisions show, click here.


There are Comments. Add yours.

 
TIP: Want to include a code block in your comment? Use <pre> or <code> tags around the desired text. Ex: <code>insert code</code>

REGISTER or login:

Forgot Password?
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy
Sort by: OldestNewest

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to: