Tech, ExplainedData platforms and AI
Product

Tape storage and archiving

Still the cheapest home for decades of data.

BEFORE ITAFTER ITTape storageand archivingQuantumCOVERSLTO cartridgesRobot libraryOffline copy30-year lifeBackup and recoverysendsolder backupsEnterprise storage systems↗ DATA CENTRESmoves rarelyread filesData kept for decadesat the lowest cost, withone copy on a shelf thatno network can reach.BEFORE ITAFTER ITTape storageand archivingQuantumCOVERSLTO cartridgesRobot libraryOffline copy30-year lifeBackupand recoverysendsolder backupsEnterprisestorage systems↗ DATA CENTRESmoves rarelyread filesData kept fordecades at thelowest cost, withone copy on ashelf that nonetwork can reach.BEFORE ITBackup and recovery↓ sends older backupsEnterprise storage systems↗ DATA CENTRES↓ moves rarely read filesTape storage and archivingQuantumLTO cartridgesRobot libraryOffline copy30-year lifeAFTER IT↓Data kept for decades at thelowest cost, with one copy on ashelf that no network can reach.
Tape archiving receives old backups and rarely read files from disk storage, keeps them on cartridges in a robot library and in a vault, and returns a file within minutes when asked.
opens a page inside this one a page in the same scene a page in another scene a concept page a label only, with no page

Tape storage and archiving keeps data on magnetic tape, which is still the cheapest place to store data that must be kept for decades and is rarely read. An archive is a store of old data, kept for reference or because a law requires it, not for daily use. Typical examples are old medical scans, broadcast video, research data and old backups.

A tape is a long strip of plastic coated with a magnetic layer, wound inside a cartridge about the size of a paperback book. Today's standard cartridge is called LTO, short for Linear Tape-Open. One LTO-9 cartridge holds 18 TB of data before compression. That is as much as a large hard disk drive, at a lower cost for each terabyte.

The cartridges sit on shelves inside a tape library. A tape library is a cabinet with a robot arm and several tape drives. When someone asks for an old file, the robot picks the right cartridge and loads it into a drive. The drive winds the tape to the right place and reads the file. This takes from one to a few minutes. A hard disk would return the same file in a fraction of a second. That delay is the trade-off.

Why is tape still used? First, it is cheap. A cartridge on a shelf uses no electricity. A disk must spin all day, and it needs power and cooling. Second, tape lasts. A cartridge is rated to keep its data for about 30 years. Third, a cartridge on a shelf is not connected to anything. So ransomware that has spread through a company's network cannot encrypt it. This makes tape a natural air-gapped backup.

Take, for example, a television news channel in Noida with 20 years of footage. It records about 2 TB of new video every day. The latest month's video stays on fast disk storage, where the editors use it. After 30 days, the archive software copies older video to two tape cartridges. One cartridge stays in the library. The other is sent to a vault in another city. Say an editor needs a clip of an election rally from 2009. She searches the archive catalogue, which lists every clip and the cartridge it is on. Within about five minutes, the clip is back on disk and she can edit it.

Tape is also the last stop for many companies' backups. The backup software first copies data to disk, so that recent files can be restored quickly. Then it moves older copies to tape for long-term keeping. Some large cloud providers also use tape for their cheapest archive tier of object storage, where a customer waits hours for a file in exchange for a very low price.

Tech, Explained is a part of ingrid.fyi. Every page explains one piece of technology in plain words.