For creative organizations, video files are typically moved from high-cost shared disk storage to archive storage after a project is completed. The archive provides secure long-term storage at a much lower cost per terabyte than the higher performance shared disk storage used for editing. Now that fast Internet connectivity is becoming more and more available, what is the better archive solution: on-premises or cloud? The answer often comes down to the cost of the alternative archive solutions. In this blog post, I explore the relative costs of on-premises LTO and cloud archive approaches.

LTO data tape cartridges in a robotic library deliver an attractive highly scalable on-premises archive solution. The cartridges provide a very stable long-term recording medium with a 30-year lifetime. And, when it comes to performance, they have high transfer rates: up to 3.2 Gb/s. The cartridges are relatively inexpensive at about $70 US for a 12 TB cartridge. However, the starting price for a networked LTO archive system is relatively high due to the initial cost of the LTO drives, the robotic library and controlling server and software.

In contrast, the starting expense for a cloud archive is low but the costs scale approximately linearly as the capacity grows. This is very different to an LTO archive which scales very cost effectively: the bigger the archive, the lower the cost per terabyte.

The price levels and pricing models from the competing cloud providers varies quite a lot. An important consideration is egress costs. This is the cost to transfer data out of a cloud ecosystem and it can be extraordinarily high. For example, the cost to download every TB from AWS or Azure is around $80 to $90. If you are using the cloud as a deep archive and are unlikely to ever need to download your files, high egress costs are okay. It also works for cloud computing applications which have minimal downloads. But for an active archive with frequent downloads, egress costs become prohibitive. The Wasabi public cloud stands out from the rest as it has competitive capacity charges and no egress costs, making it very attractive for active archives with frequent downloads. So, I will use the Wasabi pricing model for my comparison with LTO.

Let’s now crunch a few numbers and see how LTO on-prem compares with the Wasabi cloud at the 100 TB, 200 TB and 1 PB levels. Most creative organizations want to keep their video archives for a very long time and we find that typical XenData customers keep their LTO archives for about five years before they undertake some sort of system refresh. Consequently, I calculated the total cost of ownership over five years to come up with an annual cost comparison. In each case, I compare solutions that present an archive file system accessible over a network.

At the 100 TB and 200 TB levels, the LTO solution consists of a XenData X20-S Archive Appliance and a 24-slot robotic library with one LTO-8 drive used to create duplicated LTO-8 cartridges. And at the 1 PB level, the LTO solution consists of a XenData X40-S Archive Appliance and an 80-slot robotic library with two LTO-9 drives used to create duplicated LTO-9 cartridges. As for the cloud solution, the LTO systems have a file system interface.

At the 100 TB level, the Wasabi cloud is a little higher in cost but close to a comparable LTO system.

And at the 200 TB level, LTO is the clear winner with a cost about half that of cloud:

In summary, the Wasabi cloud is a cost-effective choice for video archives up to about 100 TB. But LTO becomes increasingly more attractive for higher capacity requirements. At the 200 TB level, cloud is twice the cost of on-premises LTO. And for 1 PB, cloud is more than 3 times the cost.