Several major points have emerged from our discussion of applying workloads to NAS. Consider NAS not only for cost-effective departmental workloads but also for some enterprise-level workloads that fit the NAS profile. Integrating network skills and storage skills becomes the best educational platform for effectively supporting NAS environments. Even though NAS can be justified very simply, and installed quickly with immediate results, this does not relinquish your responsibility in planning out your NAS solutions.
It should be noted that NAS, while being a cost-effective solution, could consume both manpower and additional dollar expenses as your workload begins to grow. Cost notwithstanding, the following additional considerations should be analyzed as you contemplate using NAS within departmental, remote, or data-center environments.
Departmental and Enterprise Workloads NAS configurations are easily justified for departmental settings where storage resource requirements exceed the growth of general-purpose server processing elements. In some cases, departmental applications expand beyond the scope and functionality of direct-attached solutions by their increased I/O requirements. In these cases, consider bringing these inside the data center with consideration of a mid-range or enterprise NAS configuration. Dont immediately go to SAN justification unless the total workload I/O transfer rate and user access concurrency factors demand it.
Integrating Systems and Networking Skills NAS solutions, like SANs, require an educational requirement for existing personnel. Although vendors would like to downplay this aspect, theres no getting around it. The problem is that there are few options for this type of storage networking education. As with SANs, storage expertise coupled with network expertisebest facilitates capacity planning, design, and NAS installation activities, and serves as a combined template for storage networking solutions supporting both SAN and NAS.
Plan for NAS Integration and Migration NAS allows the data center to be implemented quickly and to be safely purchased for the short term without unduly lengthening the design and planning necessary for SANs. However, this doesnt mean planning or foresight isnt necessary when it comes to NAS. NAS should be viewed in the long term more critically as it migrates through various NAS capacities and sizes with consideration for integration. Integration eventually occurs as NAS workloads work in conjunction with SAN workloads, or as NAS workloads become integrated into a SAN solution.
All the solutions within these examples follow a macro plan. The following steps are recommended when implementing NAS into production environments. As we stated with plans for SAN implementation, a macro plan is further defined with separate micro plans, tasks , and assignments. Additional steps may be required, given the level of change control managed within the data center.
Our examples demonstrate an overview of how NAS can support three common types of workloads, departmental file systems, Web read-only workloads, and specialized engineering seismic analysis application. Guidelines used in these sections can provide general direction for specific design and configuration activities.
Although not as critical for NAS, it is helpful and productive to define a test installation environment. Putting a small configuration in place provides essential first-case experiences in the configuration and operation of a NAS device. This also creates a test bed for testing future bundled model upgrades while simultaneously enabling a risk-free application testing facility.
Use the test installation to initiate a pseudo-management practice. Remote management is the most challenging activity when operating the NAS. It is also the most rapidly evolving practice, with daily changes in software tools and accepted practices. A detailed discussion of NAS management can be found in Chapter 21.
Develop a production turnover activity where a formal change window is established. In many cases, this may need to be integrated into existing change management activities within the data center. Key among these is tracking all the changes within the TCP/IP network infrastructures that support the NAS device. This can save much time when formalizing changes to the NAS configurations and integrating these into server OS and application upgrades that modify or change file-level protocols. This includes tracking key components such as the server NICs, network routers, and attached storage devices.
If you have established a formal set of production turnover and change window practices, maintaining the NAS components should become very manageable. The key area in providing maintenance to the NAS components is in recognizing the complexities of serving multiple file-level protocols. Upgrading the server OS, web software, or new storage management software may affect interactions with NAS NICs, file systems, and internal storage software.
Further establishing a maintenance matrix of NAS devices is your best defense in quelling maintenance ricochet, where upgrading or changing one component affects the operation of others. The mean time to defect recognition can be much shorter in the NAS configurations, given the fact that clients directly interact with the NAS devices. This provides an immediate feedback loop if the NAS device is down or not operating properly.
As with any active component within the network, there will be a need to monitor the operation of the NAS devices. Well cover this in more detail in the management part of the book, Part VI, but its important to note that the information gathered during this activity, regardless of product or NAS functions, plays an important role in problem management, capacity planning, and performance and fine-tuning.