Building a New Tier of Storage

The trouble with data is that not all of it is of equal value. From a storage management perspective, there are basically two types of data. There is the data in the production system and then there is everything else, which depending on what backup medium is used is generally referred to as Tier 2 or 3 storage.

The problem is data rarely falls neatly into such categories. Between all the so-called hot data in our production systems and cold data in our backup systems, there is a vast amount of data that is neither hot nor cold. This ‘warm data’ usually consists of information that someone might want to query at some point in the future, so it needs to be more accessible than backup data. Nevertheless, that information doesn’t really warrant allocation of expensive primary storage.

IBM today is trying to address this longstanding ‘warm data’ issue with the introduction of an IBM Netezza High Capacity Appliance that can store up to 10 petabytes of data in a format that can still be easily queried.

According to Razi Raziuddin, senior director of product management, the warm data issue raises its head most often in data warehousing scenarios when customers are typically trying to make data available because of compliance requirements. The IBM Netezza High Capacity Appliance is designed to balance the amount of storage typically needed to store that data against the access times needed to provide reasonable access to it.

When all is said and done there is a lot more nuance being applied these days to data management and storage systems in response to massive amounts of data that need to be managed, the cost of storing it and the actual access time that data can be reasonably queried.

Ultimately, that may add more complexity to the storage management equation than most IT organizations traditionally like, but given the chaos and costs associated with the alternative, it would seem to be the more rational alternative in a world where multiple tiers of storage are soon going to be the rule rather than the exception.