Virtualized infrastructures need smart storage strategies to achieve their full potential
With the swift deployment of virtualization in enterprise IT has come a new set of challenges to storage management. Today, getting highest value from business data requires a smarter storage management strategy—and virtualization-savvy solutions designed to make that strategy an up-and-running reality.
To understand just how completely virtualization has changed the rules of the storage game, consider the following observations from industry analysts Gartner, IDC, and ESG:
Similar complexities are reflected in the way data itself is changing. Total data volumes, for instance, are escalating so rapidly that most enterprises are struggling to keep pace. Multiplying that trend: mergers, acquisitions, and industry consolidations of all kinds that require organizations to adjust to a dramatic influx of new data as multiple IT divisions become one.
Data is also entering the enterprise in more ways, from more sources, than ever before. Examples include the billions of RFID tags currently used in asset management, mobile solutions and services, and Web 2.0 platforms. All generate important new classes of data unknown to the enterprise just a few years ago. All told, there has been an estimated tenfold increase in digital data since 2006...and the rate of increase is expected to continue.
What happens as the ongoing data tsunami continues to hit increasingly virtualized infrastructures, where in most cases data management strategies have typically not kept pace?
What happens in the case of cloud architectures, for instance, in which new virtual servers are continually created, and the data of each of those servers must be continually managed without creating a negative impact on service levels?
It's apparent that in response to emerging challenges, a modern approach to the implementation of key data management domains such as backup/restoration, disaster recovery, and data lifecycle management is essential.
In short, what organizations need today is smarter storage.
Centralized, holistic insight and command that spans multiple domains and assets
“How can IT managers get their arms around these various complexities—understanding how the total storage infrastructure is changing over time, and creating a successful strategy to manage it for highest business value? One way to answer that question lies in thinking in terms of centralized management of key storage domains.”
Pursuing smarter storage means taking into account the full array of information sources and IT assets commonly in use by the enterprise today.
Among other aspects to consider: databases, server images, file systems, storage array provisioning, tape libraries, and many others. How can IT managers get their arms around these various complexities—understanding how the total storage infrastructure is changing over time, and creating a successful strategy to manage it for highest business value?
One way to answer that question lies in thinking in terms of centralized management of key storage domains. These domains include discovery (of deployed assets and information repositories), performance management (bottleneck analysis and load balancing), configuration management (optimization and problem determination), and reporting (to track changes and emerging trends in areas such as capacity, utilization, performance, and others).
The goal should be to achieve a holistic command of them all: end-to-end network configuration, as opposed to simply management of individual components.
Toward that end, one exceptionally effective solution is IBM Tivoli Storage Productivity Center (TPC). This offering includes features and functions that align with each of those domains, giving storage administrators a central point of control over storage in as much, or as little, specific technical detail as required. Using it, they can simplify storage deployment, optimize performance and utilization, federate end-to-end management in general, and in many cases, automate everyday tasks in which efficiency and consistency are more important than direct human oversight.
Furthermore, TPC includes many virtualization-aware features that add new intelligence to how storage is managed in virtualized infrastructures. Among others:
- Discovery of elements such as virtual servers, virtual machine images, and virtual machine file systems;
- Topology and visualization, to show exactly where virtual machines exist on hosts, and reflect SAN storage as it's been allocated to different hosts and servers;
- Monitoring and reporting for physical servers and virtual guests, including health status, asset reporting, and capacity utilization;
- Problem determination and root cause analysis for storage-related problems;
- Storage provisioning that works to link any storage array to any virtual server.
In version 4.2.1, TPC includes new capabilities that deliver even more value in virtualized contexts.
For example, organizations who've made the investment in the IBM System p (IBM Power Systems minicomputer) will find that TPC can now correlate logical unit numbers (LUNs) with corresponding storage volumes, given a 1:1 mapping with the virtual I/O server.
As a result, TPC can also perform many useful tasks not possible before, such as collecting storage subsystem information and displaying volume groups, logical volumes, and files systems. Nor does this require a TPC agent to be deployed on the virtual I/O server—only TPC 4.2.1 itself, for rapid time-to-value and easy management.
And organizations with an extensive investment in VMware's virtualization portfolio will find that TPC 4.2.1 pays ample dividends in that context as well.
TPC's discovery capabilities now extend to VMware ESX server hypervisors—the basic liaison between hardware and software that drives performance for every virtual server on a host. TPC's topology and visualization capabilities can now also depict both VMware virtual machines and their locations on particular hosts, as well as storage resources used by each ESX server.
That server's properties now show up in TPC's detailed asset reports—and the same is now true for VMware guests. Finally, TPC can quantify capacity utilization for VMware environments as well, illustrating just how storage is being used (or wasted, as the case may be).
Maximize business resilience with virtualization-aware backup/restoration solutions
One of the most mission-critical aspects of storage management is undoubtedly data backup/restoration.
If data is the life blood of the organization, backup/restoration is the life insurance policy used to maximize business health by taking swift action in the event of a serious problem. It comes into play in other areas as well, such as regulation compliance—an increasingly important factor to weigh as government regulations increasingly specify how organizations should monitor and manage data throughout its complete lifecycle.
As virtualization has introduced new intelligence and flexibility into today's IT infrastructures, however, backup/restoration solutions haven't always kept up.
Older solutions and related business processes may not be well suited to current virtualized servers, in which many new complexities apply. For instance, there is the daunting possibility that if all virtual servers on a host are backed up in parallel, the host's performance will significantly decline. In such a case, the potential impact on business services driven by that host is something storage managers will need to consider carefully, and more carefully every year going forward.
Even now, most organizations still utilize traditional, in-guest backup strategies based on deployed agents within individual virtual servers. That said, there are still better and worse ways to go about these strategies.
One of the more efficient options: IBM Tivoli Storage Manager FastBack. This offering provides rapid, block-level backup of virtual server-based data across a broad range of virtual environments—not just VMware, but also KVM, Hyper-V, Solaris Containers, and others. It also supports diverse operating systems, including among others Windows, Linux, z/OS, and Solaris.
Using FastBack, organizations achieve first-rate backup performance because new data is backed up in real time, as it is created—not in a single scheduled process. FastBack is so well optimized that this can take place during production hours, yet create minimal impact on applications. Even virtual servers driving the most critical applications can typically be backed up both continually and transparently. FastBack thus helps deliver the highest levels of business resilience even though it relies on a conventional, in-guest, agent-driven architecture.
And for organizations that have largely standardized on VMware as the virtualization environment of choice, an even more advanced option is now available: IBM Tivoli Storage Manager for Virtual Environments version 6.2 (TSM for VE).
This solution works in a fundamentally different fashion from FastBack. Instead of continually backing up data using the host's own resources (such as processing power and memory), TSM for VE uses a backup server to play that role. The effect is to ensure that the host's performance is never threatened by backup processes—reducing, even more than FastBack, the odds of a potential hit on key services.
To accomplish this task, TSM for VE utilizes VMware's vStorage APIs (application programming interfaces) for data protection. Via these APIs, it accesses data directly from virtual machine storage and routes it over the network to the dedicated backup server in a single hop. Thanks to changed block tracking, incremental backups that target any new data since the last execution can run without a forced scan of the virtual server's file system. The outcome is the best available, most transparent backup process for VMware-based hosts and servers.
Also compelling: the range of recovery options, should backed-up data need to be retrieved for any reason. Data can be drawn in a single pass, and at multiple levels of granularity including files, volumes, or entire system images, to suit the business need (or the storage administrator's convenience). Besides being flexible, this restoration process is exceptionally swift—near-instant, in fact—because data becomes available while it is actually being copied in the background.
As new virtual servers are created in the VMware-based infrastructure, TSM for VE will automatically discover them and apply backup policies to them. This is truly virtualization-savvy functionality, that keeps policies mapped to virtual servers even though the total number of virtual servers may constantly be in flux and servers may even relocate across hosts unpredictably.
And because TSM for VE integrates with the centralized Tivoli Storage Manager console (versions 5.5 and above), storage managers can achieve all of these benefits in a familiar interface for effortless day-to-day administration.
Save $500 on Pulse 2012 registration!
Pulse 2012 introduces “Business Without Limits!” March 4-7 in Las Vegas, NV. Improve the economics of your business infrastructures and speed the delivery of innovative products and services. Register before December 16 and receive $500 off the standard rate for Pulse admission.