The past couple of years have been very challenging, to say the least, for many organizations. As the economy contracted, companies became financially overextended and had difficulty in meeting their debt obligations. Not surprisingly, budgets have been slashed across the board, including IT departments that are not immune to these drastic measures.
Because resources are finite even in the best of times, tradeoffs are always seen as a necessary evil when making business decisions. In tough times those perceived tradeoffs become even starker. Unless an IT project can be shown to deliver a direct, tangible benefit by either increasing revenues or lowering costs, it's unlikely to be approved in a time of budget restraint. Proposed expenditures on hardware or software infrastructure improvements can be particularly vulnerable under these circumstances. That's because the status quo is seen as "good enough," especially when shrinking business volumes no longer place a heavy strain on that infrastructure.
During tough times, priorities often change from "what's best for the company" to "what's essential for survival." Thus, projects that would likely have been approved when the economy was expanding and markets were growing often end up on the cutting room floor. The more restricted the resources, the greater the tendency is to see everything as a tradeoff. Under these circumstances, many organizations don't consider all factors, which in turn, can lead to decisions that don't deliver optimum value.
Look Beyond the Tradeoff
In The Opposable Mind, Roger Martin, dean of the Rotman School of Management at the University of Toronto, wrote, "Most integrative thinkers quite explicitly refuse to accept tradeoffs that the rest of the world tells them are unavoidable." Martin provides a number of examples of business people who, when faced with an apparent trade-off, answered, in effect, none of the above. They turned the problems on their heads so that the tradeoffs no longer applied.
In IT departments, opportunities to apply tradeoff-avoidance thinking arise frequently. What's more, these opportunities can often be seen and addressed without the need for the kind of high-level, creative, integrative thinking Martin describes in his book. Simply digging a little deeper to find all of the hidden benefits and opportunities to leverage value through innovative uses of technology may be all that is required.
Consider, for example, investments in enhanced disaster recovery (DR) technologies. Most companies already back up their data to tape nightly, but this approach has drawbacks. After a disaster, it may take several hours or, more likely, a few days to restore operations from the offsite backup tapes. Yet, even then, any data updates applied after the last backup tapes were created may be lost.
There are technologies that eliminate the shortcomings of tape-based DR, but companies often find it hard to justify their cost even in the best of times. When budget cutbacks are the order of the day, the justification becomes all the more difficult.
The conventional thinking goes something along the lines of, "Yes, we can improve our disaster recovery outcomes, and there is a value associated with that. But we'll see the return on our investment only if we experience a disaster. And disasters are exceptionally rare while the cost of installing the new technologies is certain and must be incurred up-front."
From this limited perspective, the investment in advanced DR technologies is seen as a tradeoff between a hard-dollar expenditure today versus a benefit that, if it is received at all, will be incurred only at some uncertain, later date. When capital budgets are tight, the hard-dollar expenditure often tips the balance toward forgoing the investment.
Extract Added Value
Advanced DR technologies provide a good example of how suboptimal decisions can result from incompletely assessing both sides of a perceived tradeoff's fulcrum. While their costs are typically reasonably clear, these technologies can often deliver peripheral value that may not be immediately visible. These hidden benefits may, therefore, be omitted from the decision-making process.
Consider a remote, real-time replication approach to disaster protection. Depending on the operating system platform, these technologies may go by the name of high availability (HA), although the use of the term HA varies somewhat from platform to platform. Remote, real-time replication maintains an exact replica of production servers at a remote site. This provides the ultimate in disaster recovery solutions because there is nothing to recover after a disaster, and operations can continue virtually uninterrupted. Users are simply switched to the fully redundant server at the remote site.
Evaluating these solutions based solely on the DR value they provide grossly underrates them because only unplanned downtime is taken into consideration. Although the numbers vary depending on the computing platform, unplanned downtime typically represents less than 10 percent and often less than five percent of all downtime. What's more, only about one percent of all downtime results from a disaster; i.e., an event that would result in an organization restoring its data and operations from tape-based backups. In most cases, at least 90 to 95 percent of all downtime is planned, resulting from activities such as hardware and software upgrades, database reorganizations and data backups. An HA solution that includes remote replication can eliminate virtually all of this downtime, in addition to the downtime that would otherwise result from disasters.
If one assumes that 99 percent of downtime results from something other than a disaster, traditional, tape-based DR approaches address little more than the one percent of downtime that is caused by disasters. This is because recovering a data center from tapes makes sense in only the following two circumstances:
- The online copies of production data are irrecoverably destroyed.
- The expected length of a system outage will exceed the time required to restore operations from tape. In addition, that time advantage must be sufficiently large to justify the high cost of performing a tape-based recovery.
In contrast, in a remote replication HA scenario, switchovers can typically be completed in minutes. This makes it possible to use this technology to avoid almost all downtime, both planned and unplanned.
There is tangible value to eliminating unplanned downtime, whether due to a disaster or something less catastrophic, such as a hardware crash or a temporary power failure. These events can happen at any time, including when they will do the most damage to the profitability of the business. Further, a technology that can eliminate the need for planned downtime delivers even more significant and tangible value. While companies that still operate only during traditional "business hours" can schedule downtime events during off-hours, the available maintenance window shrinks-often to zero-for organizations that have moved to 24x7 operations to support e-commerce and worldwide customers and facilities.
Search for Peripheral Uses
Furthermore, not all of the benefits that can be derived from advanced HA/DR technologies come from the value of downtime-avoidance alone. The implementation of these technologies frequently opens up peripheral opportunities that are not directly related to system and data availability issues. For example, even after deploying a remote replica server, most companies still create nightly tape backups as a last line of defense against data loss. The creation of backup tapes can be disruptive to a business that needs its systems 24x7. Some applications may have to be shut down while the backup jobs run. In addition, because backup jobs consume considerable processor resources and almost monopolize disk I/O bandwidth, the applications that can continue to run while the backup tapes are created may deliver unacceptably slow response times.
This issue disappears with the availability of a redundant backup server that contains a complete copy of all production data. The backup tapes can now be created on the replica server, without any impact on production applications. The improvement in productivity that results from better production application performance is measurable.
This option also lends itself to other uses for the replica server. Read-only processes, such as queries and batch reporting, can also be run on that system. This removes their load from the production server, thereby improving application performance and possibly deferring the need for server upgrades.
Furthermore, depending on the platform and vendor, some HA and DR products provide facilities for creating disk-based snapshots of production data. The snapshot function is generally implemented within a replica-based HA architecture. This allows snapshots to be created on the replica server, where the snapshot processes will not impact production operations. Snapshots can then be generated at will to create clean recovery points as appropriate.
The snapshot facility can also be used to create "sandboxes" containing copies of production data that come in handy when developing and testing software. Because the HA/DR software allows snapshots to be created at a push of a button, implementing these sandbox databases requires very little work on the part of the development and testing teams. Further, as the snapshots capture production data, development and testing can be performed under real-world conditions, thereby helping to minimize the risk that bugs will go undetected before implementation.
A Fresh Perspective
IT departments willing to look beyond the surface and the obvious can often eliminate apparent tradeoffs without having to choose one side or the other. When faced with a situation that appears to force a tradeoff, try to examine the problem from a different angle. Is it possible to address the issue in a way that avoids the tradeoff completely and, in doing so, delivers an uncompromisingly positive result? Or, if it is not possible to bypass the tradeoff, are there hidden benefits and opportunities on one side of the tradeoff that, if realized, would overwhelm the perceived negatives? Taking this fresh-thinking perspective can stretch the IT budget to achieve goals that one might otherwise forgo in an environment of severe financial constraints.
About the Author
As Executive Vice President and Chief Technology Officer, Alan Arnold is responsible for Vision Solutions' global technology, services and support strategy. Prior to joining Vision Solutions, Inc., he was a senior technology executive in the management consulting practice of Ernst & Young LLC. Alan is a published book author and recognized expert in advanced technologies, business process improvements and managed availability solutions. For more information, go to www.visionsolutions.com or call 800-957-4511.