One of the most mundane home cleanup tasks is re-organizing the sock drawer. Data Management has been the IT equivalent for years. It is an event that often requires someone from outside of the data center to come in and record data usage from all your various systems. Armed with this information you can make a decision on where data can be better and more cost effectively stored in your environment.
Data Management becoming cool again is driven partly by the various storage efficiency needs as we discussed in my last column, but mostly by the ability to make it a process instead of a painstaking chore. If the practice of data management can be turned into an automated process, it can run continuously and it can provide more than just a snapshot of all the information under management. Instead it can turn into a true data mining tool that not only helps with data placement but can vastly improve data discovery functions as well.
Data Management is Easier
As mentioned above, creating a data management plan used to mean hiring an outside consultant to come in once a year or so to create a series of reports that assessed an organization’s data. The problem is that these storage assessments were expensive to perform and were out of date soon after being delivered. IT teams need these storage assessments to become an automated part of the data center that are run regularly. They need to become a consistently performed process so that IT planners are always dealing with the most up to date information when making data management decisions.
Modern data assessment solutions can run either as an appliance or as a virtual machine and they can constantly scan the environment for new information. Also, thanks to increased processing capabilities and network bandwidth, these systems can analyze TBs of information in just a few minutes. They can also “dive deep” and provide content level indexing instead of file attribute basics. We call this process data profiling.
Archive Access is Easier
In years past, IT not only lacked the ability to perform an assessment of data, it also lacked the ability to efficiently take action on what was learned from that assessment. One of the big take aways from a typical storage assessment is the amount of data in the environment that had not been recently accessed and was eligible for movement to a less expensive tier of storage. The problem that IT faced was how to move this data in a way that users could still access it when it was needed.
As we discussed in our article “Unstructured Data Meet Tape Archiving Efficiency”, this problem has largely been solved thanks to network mountable archives. Both disk and tape based archives now can be accessed as easily as any other drive on the network. Users can simply be taught that if their data isn’t in their home directory, look in their archive directory. A navigation that most users could easily handle. By the same token, applications can now directly interface with these archives via that same mount point.
Data Management is once again a critical skill for data centers to possess. The ability to minimize the amount of primary storage consumption means that flash storage can be more effectively used and the more data that can be off-lined to tape, can greatly reduce a whole host of data center costs. Most importantly, much of the heavy lifting of data management can now be automated and performed almost continuously, meaning that the IT professional is making decisions based on the most up to date information. Armed with that information, those data movement decisions can now be performed as simply as a network move command; all done seamlessly to the user.