How 30 petabytes of digital archives didn’t go up in smoke
12 May 2026The digital memory we are building with our partners is growing fast. In 2022, our archive system reached 20 petabytes (PB) of digital heritage. Today, we are preserving over 30 PB of data. What challenges come with archiving digitally at such a scale? We take you behind the scenes.
Large volume, big challenges
Our content partners rely on us to preserve their digital heritage sustainably and keep every bit intact. To prevent data loss, we store three copies of each master file, spread across three locations. So in practice, that 30 PB quickly becomes a lot more.
The biggest challenge at this scale? Safely preserving this mountain of data for the future, while keeping it easily accessible today – without costs spiralling out of control.
A proven approach, even in a crisis
Last autumn, our archiving processes were put to the test. At the start of September, a smoke incident at one of our storage sites caused damage to infrastructure. The tapes that hold our material remained intact, but the drives that read and write data to tape were damaged. Read and watch how we resolved this issue.
The short version? We immediately switched to our backup systems, so our partners were still able to access and manage files. Inbound new material halted for a few weeks but was quickly resumed by renting temporary tape drives. Guaranteed continuity of service: check.
We promptly claimed insurance to replace the defective equipment. This took far longer than expected – while tape technology is our daily business, it’s largely unfamiliar territory for the insurer. In December, we finally received approval to proceed with replacements and started purchasing. Today, the new drives are fully operational and well deployed. Service fully restored: check.
What impact did this have on our content partners? Virtually none.
And most importantly: not a single bit was lost.