Every day the storage administrator of a global financial services company faces the same challenge: managing an expanding global footprint of over 200 NetApp networked-attached storage filers with an installed capacity exceeding one petabyte. Beyond the sheer magnitude of the capacity-management problem, the administrator must also satisfy increased regulatory and data retention requirements and manage within the company's resource constraints of space, servers, and power. The administrator needs transparency within this massively expanding universe and needs a way to know what storage and data he has, so he can manage it. Since 2002, this administrator has met his monitoring and reporting needs using Tek-Tools Profiler. Tek-Tools is now part of SolarWinds.
Initially he used Profiler to determine his overall CPU and disk capacity. Once that was better understood, he could analyze at the next level, and answer questions such as:
- How much of the storage pool is allocated?
- For the storage that is allocated, what is the true utilization?
- How much storage is being used for data retention?
- How much storage is snapshot related?
- Are the snapshots too large?
- How much capacity is in my Q-trees, both soft and hard?
He now uses Profiler to forecast capacity requirements for both CIFs and NFS. And he analyzes the company's usage from the top down beginning at the servers, through storage, through Q-trees, to files. Profiler also provides valuable information so that he knows how much data he has to back up.
Using Profiler to Enforce Policies
Email archives represent an area of particular risk for the company. For regulatory-compliance reasons, email archives are sensitive data. Email cannot be deleted and must be available for search. This administrator must know where the archives are stored, how many are being modified, how many are stale, and how much time it will take to back them up. With such a globally-distributed organization, it is virtually impossible to educate and control the behavior of employees to ensure that all email archives are stored in a common repository. Yet, he is responsible for ensuring that all email archives are managed in a common way, in accordance with company policy.
Using Profiler to Improve Backup Efficiency
Tek-Tools Profiler provides a powerful tool enabling the administrator to locate and migrate all PSTs to a common pool, so they can be backed up more efficiently, monitored more effectively, and managed in compliance with company policy and regulations. Distributed data is difficult to manage. Profiler’s ability to locate and migrate data enables the application of a storage service-level and uniform set of policies for each type of data. Profiler is now used to provide a deep analysis of the data. The biggest win for the company has been in understanding what is living within the volumes, classifying the data, and understanding what is driving the storage growth.
A Look Towards the Future
In the future, the administrator sees that Profiler will have an impact all the way through to the business units, which are facing additional regulatory requirements and a need for greater transparency. In his mind, everything gets more acute. Business managers and the IT managers that support them need a cleaner, better understanding of the data. They need greater reporting accuracy and the ability to control risk. Toolsets will need to understand the life of the data. Longer term, this user believes that there will eventually be an OS that monitors metadata, that understand the data, the files, and the directories.
In the meantime, users need tools, such as Profiler, to manage their storage and their data. During these difficult economic times, it is tempting to prune data, since you don’t have to store or manage what you don’t have. But in his words, “You have to know how to prune. You don’t prune with a machete.”
Action Item: Implement easy-to-install reporting and monitoring tools to analyze storage capacity, utilization, and performance. Ensure that tools provide visibility, not only to storage infrastructure, but also both top-down and drill-down views from servers, through the network, to the storage and the data. Ultimately storage is not about disk, it’s about data. Different data requires different storage and different policies. Leverage centralized monitoring and reporting tools, rather than user controls, as a more cost-effective way to enforce data policies.