Making changes in a Storage Area Network (SAN) is a daily chore for many enterprise IT administrators, but so is the risk of prolonged downtime associated with configuration errors or incompatibilities in hardware or software. A popular refrain heard from industry analysts and IT consultants is that the number one cause of downtime in the data center is due to change management errors. The thinking is that undisciplined IT changes often cause problems that result in downtime. And with the size and complexity of SANs growing - especially as enterprises deploy heterogeneous environments - the downtime risks loom even larger.

Unfortunately, too many IT administrators still employ the "plug and pray" method. That is, they go blindly into implementing SAN changes in a production environment and "pray" everything will work when it goes back online. If things go awry, they must go through a tedious, reactive process of troubleshooting to single out the mistake or incompatibility, often prolonging the downtime.

Over the next 12-18 months, consolidating server and storage resources to maximize use and lower costs will be a popular reason for implementing changes in the enterprise data center. Since making even slight changes to a SAN can be potentially disastrous from an availability and SLA perspective, IT managers need to employ the right techniques and tools to prevent the worst from happening.

The Devil Is in the Details
While it's rare among most enterprise IT organizations that any significant infrastructure upgrade isn't planned in advance, often this thinking doesn't maintain when it comes to making small adjustments or seemingly minor updates. Moreover, it usually isn't possible to understand the effects of a change in a SAN until after it's implemented.

For example, making a firmware update to a group of Host Bus Adapters could prove incompatible with the connecting SAN switches. How could IT know that would happen unless it was armed with the most current interoperability information from multiple hardware and software vendors? How can IT prevent fat-fingered errors when, for example, an IT administrator inputs the wrong port assignments to an existing SAN design? Such oversights and mistakes can be costly, often impacting the performance and availability of mission-critical applications.

The rate at which these change projects fail, therefore, is much higher than if careful planning and a means to audit the proposed changes were done in advance.

Some IT organizations farm out major changes and infrastructure upgrades to outsourcers who are paid a lot of money to do the planning and implementing. These outsourcers often charge a premium just for the detailed crosschecking of device and software compatibility, sometimes using lightweight or homegrown (though not always accurate) tools designed to simulate post-change SAN performance. Or worse, inadequate tools are sometimes used that aren't specific enough for the storage domain and instead they focus on planning the time and resources needed to complete individual tasks in a change process (i.e., Microsoft Project). This often leads to incomplete planning at a technical level, creating problems during or just after the implementation. The net result is that projects take more time to complete or fail the first time around and require a second phase to repair the problems left over from the first one.

Modeling Tools Simulate and Validate in a Safe Environment
Less than a handful of SAN change management tools are available on the market that, with varying capabilities, can help IT reduce the risk of errors and downtime associated with infrastructure changes. This class of tools can have multiple uses, even to simply ensuring that existing SANs are optimally configured for interoperability, performance, and availability. For purposes of SAN change management, these tools become powerful for not only validating that the right changes were made, but for doing frequent audits down the line to make sure the optimized environment remains unchanged.

The most compelling use for SAN change management tools lies in their ability to accurately predict outcomes. That is, the best SAN change management tools enable IT organizations to simulate how proposed changes may affect and interact with other devices and software BEFORE they're physically implemented in the SAN. Modeling functions in these tools, therefore, provide a safe environment to test different design schemes.

Good SAN change management tools also use automation to quickly upload a detailed snapshot of an entire SAN environment to save and work with as a baseline. This automation not only ensures that IT starts with an optimally configured SAN baseline, it also reduces the potential for human error and eliminates tedious manual data entry by the IT staff. This detailed topology data - which includes configuration information and device data for servers, storage devices, switches, cables, and logical access paths - can then be crosschecked against a SAN compatibility and/or configuration best-practices knowledge base to do an automated analysis of the data.

The SAN snapshot provides a baseline against which SAN changes can be modeled. By deploying SAN change management software that enables multiple scenarios to be simulated and tested, it's now possible to fully understand the holistic impact of a proposed change to the SAN before implementing a change. This critical simulation step can catch those pesky fat-fingered errors and incompatibilities. Intelligent decisions can also be made as to which change scenario carries minimal risk, while enabling the IT administrator to deliver on the required service levels for availability and performance of applications and data. Optimal change plans can then be printed out and used to support smart purchase decisions. For example, a bill of materials for the needed equipment can be generated and sent with the SAN change plan to the purchasing department along with a requisition order.

Finally, SAN change management tools accurately validate changes after they're implemented, giving IT administrators a "before" and "after" snapshot of the environment. By comparing the two snapshots against the SAN redesign plan, discrepancies can be quickly identified and corrected before the project is finalized and put into full production. This post-change verification is simply not feasible with manual entry methods and non-automated crosschecking.