Tape has experienced a resurgence as strategic and low-cost storage for massive unstructured data. And while backup remains an active use case for tape due to its value for fast site restores and anti-ransomware, tape’s future growth opportunities lie in new and emerging areas.
The tape industry is keeping pace with the storage development curve with new layers of technology. New tape technology innovations have secured tape’s position as a key technology in new data-intensive environments and emergent use cases. Advanced artificial intelligence (AI), rich applications that leverage video and other streaming data, and the Internet of Things (IoT) are shifting tape’s market drivers towards long-term retention of massive data stores.
Developments include greater intelligence like erasure coding, search, self-data integrity checking, better data management tools like remote web-based control, and more automation. Density is also improving thanks to innovations like multi-channel heads with gap scatter, and new substrates for denser recording surfaces. These improvements have positioned tape as a key technology in rich data environments.
The LTO Consortium has also been instrumental in helping to drive major advancements with a rich resource pool for multiple tape developers. This cooperation drives innovation, development, and faster go-to-market.
Emergent Usage Cases
Several industries have widely adopted AI technology for their massive data sets. Three leading users include autonomous vehicles (AV) development, media and entertainment (M&E), and video surveillance and security. All three industries generate massive data that they retain for research, reuse, and investigations. Tape lets users cost-effectively store massive amounts of active data over long time periods.
- AVs run on data, which is approaching 100 TB a day on high-end AVs. The data is both sensor information that allows the AV to react to driving conditions, and statistics on AV safety. For example, proving that an AV model is as safe as a competent human driver takes 275 million miles of driving and testing per model.
- M&E uses and reuses digital content to reach more customers and increase profits. (Sandvine’s “2018 Internet Phenomena Report” concluded that Netflix alone accounted for 15 percent of all global downstream traffic in 2017.) Streaming services easily store thousands of petabytes of compressed information, and they’re not storing that data on expensive hard disk.
- Video surveillance is evolving. Cameras generate richer video in response to advances in facial recognition and other data-intensive types of video. This does not require tape if retention periods are 30 days or less. However, some retention periods are increasing well past 30 days, especially in the intelligence and law enforcement communities. Only tape cost-effectively meets the need for retaining massive video data.
Cybersecurity Ventures calculated that 2017 ransomware damages reached $5 billion. 2018 showed no signs of slowing down, especially with the introduction of attack loop technology.
Tape is unique in data security because of its so-called air gap, simply meaning that tape cartridges are offline 95 percent of the time. In the near past, businesses could defend against ransomware simply by restoring backup over encrypted production data. However, attack loops will infect networked backup repositories too, both on the network and in the cloud. If your backup is on tape, you have an excellent chance of restoring good data.
Infrastructure / Internet of Things
Hyperscale/cloud environments are large public cloud providers offering services that require an infrastructure that has lower cost. Cloud providers may not advertise the fact, but even the largest ones frequently use tape for economical cold data storage. Another growing area is highway infrastructure investment. Both street-level sensors and AVs generate terabytes of information every day for transportation analysis.
Tape is also popular in massively scaled research environments like genomics, life sciences, and climate change. In one scenario, a major genomics sequencing lab replaced hard drives with tape. Their storage costs plummeted from $800,000 to $7,000. (Yes, that’s 800 thousand to 7 thousand dollars.) And IoT, which refers to multiple data input sources at the edge of a physical domain such as a home, office, factory, or even a smart city, is using tape storage to store the vast volumes of edge sensor data that is being used to analyze, report, and manage data to increase information efficiency.
M&E Data Mining
M&E is traditionally a large tape user because of tape’s large and inexpensive capacity for massive broadcast files. A second valuable driver for tape technology is data mining. For example, the sports industry often needs to locate video where a specific athlete hit a homerun, kicked a field goal, or delivered a gold medal-winning gymnastics performance. Or they will want to locate video of famous celebrities appearing in the stands, or monitor sponsorship commitments to make sure that their logo appears in the contractual number of programs.
Tape has always stayed competitive in areas that require long-term data retention. Today it’s becoming even more relevant because it combines durability, performance, efficient management, and low cost – making a huge difference to companies dealing with massive data and active retention.
ESG analyzed the TCO of an enterprise storage infrastructure over five years. The infrastructure included HDD on-premises arrays, hybrid cloud, and an LTO tape library. ESG reported that the tape’s TCO was nearly seven times lower than the array-based solution, and more than five times lower than the hybrid cloud.
In a fast-changing world, technology that adapts is the technology that survives. Tape continues to evolve to provide unique value for existing and emerging use cases.
Mark Pastor is director for archive solutions at Quantum. Mark represents Quantum within the Active Archive Alliance, the LTO Consortium and the Object Storage Alliance. He regularly blogs on topics relating to data protection and archival.