It is now common for public cloud services provides (CSPs) such as Amazon Web Services (AWS) and Microsoft Azure to charge data egress fees. These are monthly fees levied on top of charges for cloud services whenever a customer retrieves data from a public cloud site. For enterprise users that are now planning to increase their commitments to public cloud storage, these fees are a growing source of irritation and provide fodder to data center-based storage vendors who, when competing with public cloud storage, like to characterize is as “Hotel California.” (You can check out any time you like but you can never leave.)
The reason for users’ growing irritation is simple. Evaluator Group Research indicates that, over the next two years, enterprises will grow their capacity commitments and consequently their spending for public cloud storage by at least 20% to support a growing list of use cases for storage in a hybrid cloud environment.
Typically, enterprise users begin using public cloud storage as an adjunct to their data center storage environments by instantiation a disaster recovery capability that uses public cloud as a recovery site instead of one they must control and staff. Cloud has enabled them to implement the type of disaster recovery they’ve wanted for years but have been unable to either afford it and/or find the required IT resources. With cloud, they can instantly provision a DR site and even go so far as automating failover and failback tests on a regular basis—something that has typically been a stumbling block.
Once they’ve built the data bridge between private and public data centers, enterprise users now want to layer other use cases on top of the data bridge. These include application test and development processes, augmentation of their existing backup and recovery capabilities, and the hosting of first tier applications—all done with a storage environment that supports bi-directional data transfers. Charging egress fees for data transmitted out of the public cloud represents an ever-present toll gate that is now becoming an inhibitor to the expansion of uses cases. And there is an additional aspect of data egress charges that will become another source of irritation in multi-cloud architectures. The act of moving large amounts of data away from one CSP to be re-hosted within the enterprise data center or on another public cloud can result in a significant one-time charge which users can interpret as a lock-in strategy.
Data transfer charges leave the CSP’s business model open to disruption by someone who offers similar cloud services but without that added data toll gate. Two such CSPs are Wasabi for cloud storage and OVHCloud. OVH, for example, was founded in France in 1999 as a web hosting vendor and expanded its presence in the European market to including gaming. In May of 2017 it acquired VMware’s vCloud Air public cloud business including seven data centers and expanded its operations to include the US. Today, OVH serves 1.5M customers from 28 worldwide data center on 4 continents staffed by over 2,500 employees. It builds its own infrastructure along the lines of other web-scale service providers such as Google.
OVH has typically appealed to two customer types: small to medium scale businesses (SMBs) and enterprise users needing a scalable, automated cloud environment for developing, hosting and re-hosting applications; and vendors of SaaS and other web-facing applications. Enterprise users could be attracted simply on the basis of expanding their hybrid clouds in OVH’s direction. But enterprises are now also using cloud services to host their own SaaS and web-facing applications for customers, clients, patients (in healthcare) and partners. The fact that OVH does not charge for ingress or egress charges makes its cloud particularly appealing to these dual-purpose cloud users as well.
