This article was contributed to VentureApp by Stefano Maffulli, director of cloud marketing & community at DreamHost, a cloud and hosting company which provides domain registration, web hosting and cloud services to 1.5 million sites, blogs and applications, and supports over 400,000 web designers, developers, content creators, small businesses and entrepreneurs.
When I started using software, the first question I learned to ask was: how much will it cost me to migrate to other software? The same question applies to clouds: how hard will it be to leave it? What happens when my business needs change and it makes more sense to keep the cloud in-house? It may seem counterintuitive to start from the end, but experience taught me that it makes perfect sense.
When I was a young architect in the early 90s, I heard the story of an architecture firm who modernized their offices buying software and hardware from a great company: they had top-of-the-line equipment, fantastic workstations, 3D capable and hyper-modern CAD software. Everything was great — until the proprietary model was disrupted by cheap workstations running commodity hardware. Over five years worth of drawings, including drafts and final projects, were locked into hardware and software that nobody could support anymore, stored in an incompatible format on 5-year-old computers that could die at any time. I still feel the pain of that company to migrate from that legacy system. My love for open source started around then: I never wanted to lose control of my means of production.
The lesson of the architects translates to cloud providers: isn’t “the cloud” just someone else’s computer? Before deciding which computers deserve to keep your business applications and data online, these questions are worth asking. (Spoiler alert: almost all of the answers lead to clouds based on OpenStack, a leader in open source cloud solutions.)
- How much will it cost me to operate a virtual infrastructure? Not owning computers and running applications using on-demand means shifting costs from capital expenses to operating expenses. Startups can avoid spending money to buy computing capacities today based on future growth and use pay-as-you-go infrastructure instead. Most public cloud providers offer hourly prices for computing power, storage and bandwidth consumption. Comparing offers from different cloud providers can be very hard: it’s good to run proof-of-concepts with two or three providers and compare costs based on these examples.
- What standard tools and libraries are available? Big proprietary public clouds like Amazon AWS have spawned an enormous ecosystem that small cloud providers based on proprietary code cannot beat. OpenStack ecosystem is gigantic, too: all cloud providers built on OpenStack can count on the enormity of resources invested by IT giants like Paypal, Walmart, IBM, HP, Intel, AT&T. Small cloud providers like DreamHost are able to offer a credible alternative to the giant proprietary public cloud providers thanks to OpenStack.
- What talent pool will I be able to attract to operate this cloud? The largest expense in a startup’s budget is people. Picking an obscure technology for your IT needs will probably make it harder to recruit people as you grow. Sticking to widely adopted cloud technologies will help your funding last a little longer. Sticking to clouds based on OpenStack is a safe bet, with a growing pool of people trained to use OpenStack.
- Where is this cloud located? Despite the technical innovations, legal and social implications make physical proximity an important criteria for choosing a cloud. Moving data across public internet is still affected by physical proximity. If your customers are mostly in the U.S., picking clouds that have data centers in the U.S. is a safe choice. Norms and regulations in many countries mandate that data is kept within national borders. And support in local language and timezone makes things a lot easier when problems arise. OpenStack members have public clouds available almost on every continent.
- Finally, how easily will I be able to move the cloud technology in-house or to another cloud? This question is often heard when startups reach a considerable size. Dropbox, for example, was using Amazon S3 service to store files. Everything went smoothly — until the service became very popular. Now that Dropbox has millions of dollars in revenue, they decided to migrate off of Amazon’s cloud to keep more money in their pockets. Similarly, successful startups may decide to move their applications to a cheaper/better public cloud or create their own. Picking an OpenStack provider allows both options easily, with lower costs than, say, trying to replicate Amazon AWS or Google Cloud.
All things considered, choosing open source solutions can be a safe bet and asking these important questions in early stages can make or break a new venture. The public clouds powered by OpenStack are designed to buttress your future endeavors.
Want more content like this delivered to your inbox? Join 10,000 people from companies like HubSpot, Drizly, and DraftKings that receive our newsletter once a week.