Dynamic Infrastructure
Standard traffic pattern for a startup’s infrastructure is generally defined by figure 1. It can be crudely represented as a flat line with very well defined usage spike that correlate with planned or unplanned (also called ‘Black Swan‘) events. Those events are crucial to the development of your startup: they generate traction allowing your business to launch successfully.
Untimely system failure during those events can seriously limit your startup development. Some mitigation is possible with a carefully defined capacity forecast, however, even the best manager can’t plan for the unexpected.
How to survive such events ?
Dynamic infrastructure is the answer Les Laboratoires Phoenix give to this problem. Using very well known free software (GPL license), the infrastructure is made aware of a series of very specific metrics on which triggers are built. Such infrastructure is able to self-heal, automatically scale out to mitigate ‘Black Swan Events’ or even scale in to lower cost.
The scaling aspect can be made with an hybrid system of dedicated servers and cloud computing providers. Such setup can greatly reduce operation cost for clients because they are only billed for used computing power (/hrs).