;
If you are a user of LTO tape technology then you know that it has outstanding reliability. How much LTO tape is being used? Well, let’s put it this way, in the last two years over 250 exabytes of total LTO tape capacity (compressed) was shipped to users around the world. Whew, that’s a lot! You may not know all of the key factors that contribute to LTO technologies excellent reliability but if you are a little curious we are going to reveal some of what’s behind the data integrity curtain in this BlogBytes article. Let’s take a look.
“Excellence, then, is not an act but a habit,” Aristotle. You could say that the design and development of LTO technology is an excellent habit as it has evolved over the last 20 years progressing to LTO generation-9. LTO-9 technology boasts up to 45 TBs of compressed capacity per cartridge (18 TBs native) which is a50% increase in capacityover LTO-8, and a 1400% increase over LTO-5technology that debuted a decade ago. LTO-9 technology boasts the fastest LTO drives to date with up to 1,000 MB per second drive data transfer rate compressed, or up to 3.6 TB per hour, per drive, to help handle the most challenging backup and archive workloads. That’s not all, in the data reliability category LTO-9 is better than 1 uncorrectable error event in 1020 user bits which is typically referred to as uncorrectable bit error rate or UBER, translating to at least 17 nines of durability. Let’s break that down a little bit.
What is an uncorrectable bit error (UBER)? It is a latent hard read error – a key data reliability metric for all data storage devices including HDD, SSD, and Tape. As noted in a recent LTO-9 Technology and User Data Reliability Analysis paper, “ … due to LTO’s unique format which is based on orthogonal interleaved 2-dimensional 32 channel Reed Solomon error correction codes, the probability of an UBER event is orders of magnitude lower than HDD.” How many orders of magnitude? According to the paper, an LTO-9 cartridge with 18 TB capacity is about 5 orders of magnitude better than an 18 TB HDD. To describe it another way, the HDD example would experience an UBER about every 125 terabytes or about every 7 HDDs. LTO-9 technology would only experience an UBER about every 12.5 zettabytes which is 12.5 billion terabytes or nearly 700 million LTO-9 cartridges!
How does LTO technology help prevent these types of errors from going undetected? There are a number of data integrity mechanisms. For example, LTO technology can help fully verify the data write process; “LTO has built in resilience to write mode latent errors (UBER) due to its Read-While-Write architecture where such write-related errors are detected and rewritten during writing.” This means that LTO technology reads the data right after it writes it to double check its accuracy. If it is not accurate it
re-writes the data accurately. Not only is LTO-9 technologies UBER five orders of magnitude over HDD but it is also one order of magnitude over LTO-8 technology. It just keeps getting better!
There is a lot more that can be discovered about the data integrity characteristics and robustness of LTO-9 technology including discussions about the tape media and drive head mechanism. For a deeper dive on this subject please see the LTO-9 Technology and User Reliability Analysis paper here.