Cloud Storage = Somebody Else’s Computer
What is the benefit of cloud for large scale storage of, e.g., a multi petabyte archive? This may shock you:
Public Cloud cost for 1PB (priced at $.024 per GB) for 5 years = $1,440,000 + likely very large egress fees. You can up to double that if you want to keep your data with two companies for safety.
Hardware / software / support cost of on-premises 1PB on spinning disk private cloud is less than 1/3rd of the cost over the period.
So, is it worth $1m to let someone else hold your data? Definitely “no”. But maybe “yes”. For one thing, it depends on the workflow. For instance, if your concern is distributing your data around the world or if you do not wish to manage LTO migrations of rarely accessed content then cloud might be the way forwards.
But it’s great to see some of the hyperbole of “cloud” dissipate – it is not the answer to everything.
What is changing in the stratosphere of cloud is the number of integrations into that platform from people wanting a piece of the pie. E.g., hook a cloud video analytics engine to your archive and get a bunch of results. It’s neat and easy but it is expensive and at times of very dubious quality.