I ordered the first development server for Project Wolf today. Choosing a server wasn’t as easy at I thought it would be, as there were many variables to consider. However at the end of the day, after much debating amongst the team, I decided on a local server.
We could have gone with some cheap overseas hosting, but this project is top notch and should be treated as such. Below are some of the points we discussed before deciding on choosing a development server.
- Cost Effectiveness
Without a doubt, hosting locally is drastically more expensive when compared to hosting overseas. Not in terms of hardware but bandwidth costs. You are looking to pay anywhere from R93 to R123 per GB (Gigabyte). However, the options to follow justify the local purchase. - Operations
If you look at it from an operational point of view, the majority of our first clients will be South African. We want to be able to deliver true local speed. Why waste time moving from an overseas server to a local server? In demonstration scenarios, you want the service to load as fast as possible, even on a really slow connection. That being said, we will be able to quickly and easily add capacity for the Americas, Europe and Asia as we grow. - Testing, Implementation and Deployment
Having a local server makes life easier and faster. Every single file upload and page refresh happens much faster, add all the seconds up and we save a lot of production time. - IP Protection and Security
The great thing about having a local server, that you are in control of, is that you have legal recourse on SLA (Service Level Agreements), which means it’s easier to protect your service. Another thing I wasn’t happy about was having all our code on a foreign server, which does back-ups to other foreign servers. I want to sleep at night knowing I am in control of our data and that it’s only a drive away.
Tags: Development, Mobile, project wolf, projects, web, wolf, younique
Your decision is going cost you a lot of money. Why not start with offshore hosting in Europe, at a low operating cost, and as your service proves it’s viability start worrying about improving local access speeds?
Seems silly to risk the viability of a start-up for the romantic idea that people care about the 4s extra your page loads will take.
* Costs: SA is probably between 10 and 1000 times more expensive than hosting overseas
* Operations: Throw a development and testing server on the end of a 4mbps DSL, this can also double as a backup and code safe.
* Testing: see point above
* IP/SLA security: SA has pretty much no “protection” for code (which hopefully never changes), so your point about IP is moot.
Being able to have recourse for your SLA means you lose out on the service level a company aiming at a _global_ market offers.
My personal opinions may be influenced by working for a company which hosts offshore because we have more than half a million south african clients.
@Darb: You’re points are all extremely valid, the rest of the team raised the exact same points. However, due to the nature of the product, speed is an essential and time-to-market is another important factor, therefore we had to make the decision.
In terms of recourse for overseas SLA, we don’t have the time nor the product right now, but once we do, we will have some great overseas partners which we won’t have to worry about.
Thnx for sharing, good luck with that, sure you will make a success!
You might be interested in this:
http://www.swimgeek.com/blog/2008/08/13/teraco-is-open-for-business/