What is driving new backup tape library adoption today?
What we are seeing today is pretty much a concentration of libraries in the higher end. Smaller enterprises are generally opting for disk and cloud-based storage for backup. However, the big data movement is making even the smallest enterprise reconsider tape.
Big data, particularly when discussing in the context of popular NoSQL processing such as Hadoop, is actually optimized around streaming, which makes tape particularly suitable when used in conjunction with disk. Also, an on-site backup tape library used for long-term (or active) archiving is faster than retrieving data from the cloud.
The biggest driver to new adoption of tape is that tape (and, by extension, tape libraries) is no longer just for backup. Actually, it never was! Disk is simply too expensive for archiving both in terms of up-front capital costs, but also operational costs (such as power, cooling, floor space, etc.). Tape is appealing due to its density, relatively low operational costs and its sequential access works well with contemporary data analysis methods.
This was first published in January 2014