The durability, low cost and high density of tape make it the ideal media for handling extremely large amounts of information. But tapes also come with one significant drawback: SPEED.
Automated tape libraries help a bit by automating the task of finding and loading the right tape faster than a human could ever do it. But it still wasn’t enough. In order to use the data on any given tape, you must first locate it, load it into the tape drive, and then read it sequentially until you find the file you were looking for.
For a single tape, this is slightly inconvenient. But for a large tape library, it’s completely impractical.
In order to work effectively, automated tape libraries needed a significant boost in speed to take them to the next level. This is why Virtual Tape Libraries or VTLs were created.
Instead of reading/writing directly to tape, VTLs, would keep a large disk-based buffer of Virtual tapes which could be quickly written/read. Often times, this buffer could represent as many as 256 or 512 tapes.
This approach provided 2 significant benefits:
Virtual copies of tapes could be instantly accessed in the buffer without having to be loaded into the reader. Now you could access 20 different virtual tapes in the same time it would take to find, load and read just 1 or 2 tapes.
With simple automated tape libraries, storage devices were often under-used. To save a file, you’d pick up a tape, write to it, and then put it back. Now, you could load up the entire virtual tape before putting it back. Or better yet, find a way to optimize file storage across many virtual devices. For example, several smaller virtual tapes could be copied to one physical tape in order to save space.
When you think about how VTLs streamline the data management process, it’s almost hard to imagine that all of this work was once done by hand not too long ago. If it wasn’t for the speed and efficiency offered by these devices, many data-intensive industries simply could not exist.