What seems to be the problem? Information lifecycle management (ILM) as a concept offers the potential to manage more efficiently both your data and the storage on which it resides. For example, data that does not need the high performance of Fibre Channel disks can be put on less expensive capacity-rich SATA disks. Moreover, information as it changes value over its lifecycle probably does not require as high a level of availability, which can translate into a lower cost to protect it. However, figuring out how to put in place an effective process and strategy for ILM has been problematic at times.



Another approach called data lifecycle management or hierarchical storage management (HSM) is sometimes taken. This approach moves individual blocks of data between tiers of storage rather than dealing with blocks of data that are organized into files or relational database structures. That seems simple enough and can provide for a form of tiered storage, but beware! This approach is very limiting and not a substitute for ILM.



What do you need to know? The original approach to moving data from one tier of storage to another using blocks was called hierarchical storage management (HSM). HSM was originally used to move blocks of data from mainframe disks to tape media. HSM works on the basis of block metadata. Block metadata is very simplistic. Typically, if the date of creation or the date of access exceeds a threshold, the data block is migrated. And sometimes the term data lifecycle management is used that does the same thing.



Those of you are on the shady side of age 30 can remember when disk storage was very expensive (i.e., tens of dollars or more per megabyte). Tape was a lot less expensive. Consequently, the original use of HSM to move probably no longer useful data to less expensive tape media was important to mainframe shops. However, today the cost of disks is typically a less important cost consideration than the software and people intelligence that is needed to manage the data. And those intelligences cannot operate effectively simply on the age of the data or the last time accessed as that does not give enough information for decision-making. Intelligence must be able to work on at least the file metadata. For example, should particular files be moved to a different tier of storage that can be managed as compliance data? Should some data receive extra data protection by being remotely mirrored to a remote site in addition to standard backup processes while the standard backup process are all that needs to be done for other data? Blindly using block metadata is not enough.



What can you do about it? Now please note that the terms HSM and data lifecycle management are very respectable terms. Newer HSM solutions could very well deal with file metadata say rather than blocks, as could a data lifecycle management solution. The key is not the label, but rather the functionality.



You can examine very closely any proposal that promises to help you tier storage. How do you migrate data between tiers of storage? If age is the only criteria, the approach is probably just the movement of blocks. If that is all that you want now and in the foreseeable future fine. However, that is very limiting.



Instead, turn your attention to real ILM. Information lifecycle management is the policy-driven management of information as it changes value through the full range of its lifecycle from conception to disposition. If you want the benefits of ILM, you will have a lot more work to do (such as data classification). As the old saying goes, no gain without pain. However, you may have even more pain if you have to figure out how to wiggle out of what appeared to be a simple block-migration solution, but one that made it hard to employ a more benefit-rich solution that helps out both compliance and data protection.

For ILM, you had best not accept any simple substitutes or imitations that work only on blocks. That apparent simple solution is unlikely to be even adequate.

ITworld.com

• Email
• Print
• Share
  • Slashdot
  • Digg
  • Stumble
  • Reddit
  • Newsvine