Today is the day of a remarkable new entry into the Canadian wireless space, the launch of WIND Mobile. It is a launch I’m incredibly excited about, but that in itself shall be left to another post. The launch of WIND, just as with virtually any other major product/company launch, was not without problems. Mere minutes after I first visited the new online store portal http://shop.windmobile.ca, the connection was cut, and I was no longer able to access the online shop. It would seem that the problem was simply due to immense demand on the servers. It’s quite a common experience for hot new product/service launches. Just ask Microsoft or Apple.
Anyway, this started a train of thoughts. Were I planning the technology end of a product launch, how would I prepare to deal with unanticipated demand? The problem with unanticipated demand is that you don’t know how many resources need to be set aside to cope with or load-balance. To me the answer is relatively simple. Leverage cloud computing resources. Build and use your front end for launch day on the cloud, preferably a cloud service that auto-scales to meet demand. Microsoft Azure would seem a great candidate for such a task, as it seems to extend the Windows Server platform nicely. You pay for the extra resources and bandwidth you utilize, but having a working front end on the most important day in your product’s lifecycle far outweighs the cost.
Simply planning ahead with the thought to use the ‘unlimited’ resources of a cloud computing platform for a product launch would be my preferred method of preparing for a major product launch.
Please leave me your thoughts below.
Related posts:
- Cloud Computing & IT
- Microsoft vs. iPad
- Rogers & DOCSIS 3.0
- On the Security of Client Platforms
- VMWare ESXi as a Media Server
Tags: Cloud, cloud computing, demand, launch day, load balancing, product launch, Server, WIND
Wes, great thoughts. Companies should definitely be considering cloud solutions in their platforms, – especially for auto scaling on demand, I agree. Cloud solutions have been around for quite a while now, and a company like WIND would have been smart to take advantage of them for their launch. They can’t afford to be perceived as behind the times, unreliable, or unable to meet demand by prospective customers looking to make a change.
Amazon’s EC2 cloud is perfect for this kind of demand and has been around far longer than Azure too. Plus, you’re able to run a stable, scalable, server ready OS on the Amazon machines also, which is great. This would have definitely solved WIND’s problems this morning – oh well, perhaps next time Globalive launches in another country they’ll think ahead!
Agreed John. I personally generally prefer Amazon EC2 over Microsoft Azure, but they fundamentally are different. EC2 is designed as a standalone computing platform, where Azure is really designed as a complimentary service to an internal infrastructure. I think it setting up Azure as an extender to an internal infrastructure would be much more feasible than setting up new Windows Server boxes on the Amazon EC2 computing platform and dealing with general Microsoft Licensing issues. The licenses for setting up a launch day portal also are not as flexible to maintain.