Tape storage has been around since more than 60 years now. Since 1952, to be exact, when IBM invented the concept of magnetic tape storage. There have been many great inventions and improvements in that time frame, and although the statement "Tape is dead!" was exclaimed quite regularly over the last two decades, tape storage is still here! And I believe it's here to stay, at least for as far into the future as I dare to look.
Let me explain, why:
Nobody in the storage industry would deny that tape still has a significant market share, but lost it's predominant position as primary backup solution. Nowadays, disks are (fairly) cheap, disks are (fairly) fast, and disks have (fairly) large capacity, and that's why they are the backup medium of choice for the majority.
But nevertheless, in the past couple of months or so, I noticed more news and discussions about tape storage than usual. Let me share some of them:
First, of course, there's the regular (and usual) drum roll on new hardware releases. More capacity, longer tape medium, higher track density, etc.
But isn't this just "more of the same"? Technology leaps (or creeps?) forward, so it's just a matter of time until your recently "leading edge" device turns into "old iron".
I'll leave it to competitive analysts to compare facts and figures (whether it's performance, reliability, energy consumption etc.) of these announcements!
Then, with a more strategic (and non-proprietary, though IBM-focused) view, IDC puts a "Technology Spotlight" on LTFS and tape usage strategy. Especially in the media and entertainment (M&E) industry, tape tends to be the primary media type, and is expected to even grow it's position as "a foundational tier in long-term data preservation archive solutions".
LTFS, to step away from the pure hardware layer, is designed to integrate into the scalable data management of General Parallel File System (GPFS), adding another economical and technological value-add to data center operation. This might be the main reason for the fact that not M&E, but the IT Services industry is the main adopter of LTFS technology.
In their paper, IDC conclude that "LTFS technology — in particular IBM's LTFS Enterprise Edition — can be the foundation of an affordable and effective archiving strategy". And isn't that what we all all looking for?
Just recently, the Tape Storage Council (yes, I'm aware they are biased) released a Memo, titled "Tape Storage Meets Long-Term Data Retention Challenges" (you can also view the related webinar and panel discussion on YouTube). Let me summarize it's message in three sentences:
There's no chance to avoid the buzzwords of "Cloud", "Big Data" and "Compliance", when explaining that the current trend is to ‘save everything, forever’. And here, who would doubt it, lies one of the core benefits of tape storage: It's by far the most cost-effective storage technology for long-term data retention!
And last, but not least, David Floyner took even one step further by announcing the "Rise of the Machines: The Rebirth of Tape" on Wikibon!
He, too, sees LTFS and the increased tape hardware performance (especially in comparison with the stagnating HDD IOPS performance) as driving factors for the new momentum of tape storage.
But he adds another twist to the discussion: Flash storage!
The combination of (extremely fast) Flash storage and (low-cost, long-term, high-density and high-bandwidth) LTFS tape technology might make much of today's traditional disk technology obsolete over the next decade! Remember the term "Flape", which has been shaped by Wikibon for this concept. You might hear a lot more about it in the near future...
This being said, I hope you understand why I'm so enthusiastic about being in the tape storage business...
... and I have not even started to talk about hierarchical storage management yet - I'll leave this as a topic for another blog article!