What is the Point of LTO?
By Jon Morgan, CEO, Object Matrix
LTO is under attack from a wide variety of new technologies that challenge the very existence of the format and its relevance. Coupled with the fact that many broadcasters are pushing more of their archive content onto local or private cloud platforms, it is hard to see how LTO is still standing.
For sports broadcasters, it is more important than ever to be able to see the content you have then consolidate it to create a meaningful archive that can be accessed at any time for replays or clips from old matches. And furthermore there is a strong desire to re-archive older tape formats before those tapes become unreadable and sports events of the past become completely lost. Is it relevant to put those old tapes back on to another tape format?
Is there even a point to the tape format anymore or is it only surviving because of “better the devil you know”; doomed to disappear altogether once people become educated to alternatives?
The demand for online sports
Many stats point to a shake-up in traditional TV for sports coverage. Take the NFL, for example, where, according to Nielsen, viewing is down 8.2% from the same period last year. And, last year Sky in the UK stated a strong shift of viewing of the Premier League to their online platforms. Amazon are rumoured to be bidding to show future Premier League seasons. Undoubtedly, the shift is set to continue, as younger viewers in particular are more likely to watch sports on digital platforms.
For digital viewing instant access to content is a must. No wonder broadcasters are pushing more content onto on-premises or off-premises (private or public) cloud platforms. The benefits can be fantastic: with instant access to catch-up reels and to pull out historic content. What other type of content suddenly prompts the need to find something that was aired several decades ago? But that happens all the time in sports, whether that be a rundown of all the tries scored by a retiring rugby player, the re-run of a historic win, or the moment a team was promoted or even demoted.
Sports broadcasters need superfast access, and not just the low-resolution proxies, to make effective use of these moments. At a recent Amazon seminar in LA a number of speakers, from global broadcasters, explained how the need for instant access meant they were hammering their LTO libraries in such a way that failure rates inevitably rose dramatically. They determined that for frequent, fast access LTO is not the appropriate platform.
Or course, tape used to be the only real option for media libraries. However, now the cost of both disk and cloud services have come down a great deal, which makes me wonder whether anyone actually prefers LTO to disk for an active archive.
Tape will live on
Whilst it is true that new formats are being released all the time and tape can’t efficiently do all that disk storage can do, it isn’t dead yet. Not only that, I believe it will live on for the foreseeable future. Given the low costs of the tape media itself, it is easy to see why it has a place for “fire and forget” workflows. Of course, there are a lot of those type of workflows, even in the sports arena. Let’s face it, we all know there are matches we never want to watch again. Even sports entertainment eventually has a shelf live, and being able to literally put a tape on a shelf with a low dollar per gigabyte cost makes a lot of sense for that content.
LTO can also integrate quite nicely with other storage solutions. For our part, our solution is used in a number of deployments with LTO partners, where MatrixStore is an active cache and the LTO is used for the “dormant” assets.
The Argument Will Continue
The argument will likely continue over coming years, with some stating content is more secure on tape, whilst others argue the opposite. The argument around cost is also a hotly debated one, with many conflicting studies on whether or not the cost gap is closing.
Ultimately, I believe a number of things:
1. Tape will be around for many years to come, however it will be used in fewer use cases than today.
One example is that it might have been previously been ok to keep old footage on the shelf just-in-case it is needed, now, old footage is sold to sports teams and is looked at for analytics and AI metadata extraction. Assuming that footage won’t be re-used at all after the live event is a dead concept. However, it could be that there is footage from minor sports events that after a certain period of time is deemed to have a low probability of being looked at again.
Maybe those use cases will be cause for a lot of tapes to be sold making the tape industry appear “steady” but only in the context of a world that is storing increasing amounts of data.
2. Tape will only really be chosen on the basis of financial reasons, as there are very few valid technical reasons to consider it.
The pure fact that tape has to wind its way through to a file, over disk, which can get to the file in nanoseconds is the clear and simple problem. Add to that problems of maintaining tapes, problems of migration between formats and problems of tape versions going out of date and it is a technology that a few love but many hate.
3. More sports broadcasters will be looking for the value than an active archive can bring.
An active archive adds value to an organisation, making it much easier for broadcasters to monetise content. This is because it is much easier to find and extract at any given moment and it also enables automated workflows which make the whole process much more efficient. An archive at rest cannot add value in this way and purely adds cost. For example, a Swedish minor league football broadcaster now sells the footage of the games played by their rivals to the clubs for the purpose of analytics. Furthermore, old games can be purchased for viewing by fans. Although the turnover for lower league teams is fairly low it still funds the broadcasting and archiving operations.
Essentially, the right storage really depends on what your priorities are. If you need deep cold storage, LTO or ODA may well be less expensive. However, if you need instant or frequent access to your content, then LTO simply won’t cut it.
what a long post.
You go out and buy a new HP Ultrium LTO 8. Please tell me how you read an LTO4 or LTO5 tape with your new LTO8 drive ? (let's say your old LTO5 product died).
Rescue 1, Inc.
After seeing $300K-500K price quote from various service providers for medium-large data to be stored and on the cloud - particularly AWS, I don't think that's profitable unless the ad or revenue is so high that $500K is barely making a small dent to the revenue. I rather have my own in-house large scale HP LTO-8 tape library than throwing away half a million USD$ every year to AWS.
LTO seems like a oligopoly to keep IBM, Quantum and HP in full control and to prevent competition. Sony used to have the AIT which I have seen working beautifully with FCP7, though a pity it didn't caught on. In LTO case, as a tape medium, I can never understand why the drives should be so bloody expensive even after so many years.
I just hope somebody will write a backup software that can write to 50 Gb Blu-ray using LTFS one day.
[Melvin Chong] "I just hope somebody will write a backup software that can write to 50 Gb Blu-ray using LTFS one day."
Pardon! Surely you mean like the packet writing software we used to use on CD's and DVD's last century? Where you just insert the disk and format it in a special way and then you read/write to it as if it was a USB Stick? Why can't you use the packet writing software that has been updated for Blu-rays? Here's one example where it mentions this if you scroll down to the bottom of the page.
On page 25 of this document Nero's InCd program formats BD-R's using UDF 2.6 and BD-RE's using UDF 2.5 so this is another program that supports packet writing.
I'm reasonably certain there are a few other manufacturers of packet writing software available and if not you have at least these two currently available. Kindly note that deleted space is not recoverable on regular write once Blu-rays but should be on re-writable Blu-ray disks BD-RE's. If you have this available you don't really need LTFS for disks, however, recipients of these disks will probably need their own copy of the packet writing software unless the companies release free disk readers for other people to read them and Nero usually gives away a free disk reader by way of download,
LTFS is widely adopted by the industry and if such software exists it could work well with many media asset management auto- tiering solutions.
There seems to be a number of opinions in this thread and plenty of room for argument. I recall the articles back in the late 1990's and early 2000's where disk manufactures claimed "Tape is Dead". They were clearly wrong. Tape is not dead. So why would a cloud provider promote using tape? While LTFS allows tape to act like a random access device, it is not a replacement for them. LTFS technology is not intended as a backup solution, but rather an archive solution that is easy to use. The whole point of archiving data is that it is not used often, low cost storage and secure. The cost factor is significantly lower than disk or flash. Businesses use technology to save money not just because it's "cool" or it's the latest and greatest.