Server and network virtualization pose complexities to network change and configuration management (NCCM). But implementing change management in storage area networks (SANs) is the mother of all challenges.
That's because mapping, documenting and monitoring SANs is extremely complex. SANs centralize enterprise storage by forming a network between all of the storage devices. The SAN then connects with the data center and on to the rest of the network.
Change management for the SAN means mapping and documenting complex relationships between hundreds (or more) of interconnected servers, as well as switches and storage arrays. SAN change management also involves the intricacies of documenting the relationships between virtual machines and their hosts. Beyond the components, monitoring SANs requires tracking thousands of access paths in real time. In a SAN with only 50 servers, tens of thousands of paths must be documented and monitored.
To add to the complexity, network managers create zones on SAN switches so that servers see only specified applications. This is done for both security and organizational purposes. Also, managers assign storage volumes to specific applications and they must document every change made to any of these configurations.
SAN change management is crucial since application availability relies on functioning storage networks. Approximately 50% of problems in the SAN are related to change.
Ultimately, SAN change and configuration management tools aim to avoid application outages and maintain service-level agreements by managing change implementation more efficiently.
What can SAN change and configuration management tools do?
SAN change management tools generally enable automated component and configuration discovery, mapping, centralized control and reporting. The software also monitors and reports on paths, as well as the health of business processes and applications. Reporting enables network managers to use this information to make decisions about changes they can make without disturbing existing configurations.
While SAN change management applications often fall under the umbrella of SAN provisioning and capacity planning tools, the software must also integrate into existing network change and configuration management. Many SAN change management applications integrate into existing server and network change management databases (CMDBs), offering a single-pane view of the entire network from the core to storage.
Getting to the root-cause with SAN change management tools
While there are similarities in network, server and SAN change management applications, SAN-specific tools often have more complex functions. One of those functions is root-cause analysis.
Root-cause analysis uses information collected either from real-time monitoring or from snapshots of data paths to pinpoint the cause of a problem. The goal is to first create a baseline of what a healthy configuration should look like. Then, when a problem occurs,
With root-cause analysis, managers use information collected to create algorithms for fixing problems. Automated root-cause analysis is one way for storage managers to make sense of the large influx of information that comes in from real-time monitoring alerts.
Predicting the future with SAN change management tools
Some tools also include predictive applications. They model the SAN and then simulate change in order to measure the potential impact on the system.
Some of these applications create simulation in an out-of-band network. Most storage devices include common information model (CIM) agents that managers can use to create a model SAN. Using this model, they can introduce new change and also replay incidents that occurred to examine alternative troubleshooting methods.
ITIL and SAN change management policy
IT Infrastructure Library (ITIL) -- the internationally recognized set of policies and procedures for managing IT systems -- briefly addresses storage change management. While ITIL is meant to offer a standardized approach to everything from cabling infrastructure to computer installation, the latest version focuses heavily on service management. Because of that, there is a section that addresses storage archiving and policy specifically. ITIL calls for a dedicated team to manage storage. It also outlines how, where and for how long data should be stored.
For more background on network performance management in general, visit SearchNetworking.com's network performance management topic page.
This was first published in September 2009