Dynamically Scaling Resources: How to Save 60% on Costs with Cloud Hosting for Allegro
By Tandy Lofland, Mark Swann, Jason Ewton & Charlie Palmer
Reducing Cloud Hosting Costs for Allegro with Dynamically Scaling ResourcesWith the increasing popularity and availability of cloud computing in today’s business marketplace, many companies are considering cloud hosted solutions as opposed to the more traditional option of managing servers in-house. Coupled with high reliability, a cloud server grid offers the benefit of having complete control over creating, starting, stopping, and scaling servers instantly while only paying for the servers when they are running. By utilizing this functionality in an Allegro implementation, a business can drastically increase server grid performance while minimizing server uptime, thereby greatly, reducing the overall cost of hosting. At one of Opportune’s retail power clients, costs were reduced by 60%!
Over the course of each day and week, there will be distinct periods of high and low system resource usage in any ETRM implementation. Traditionally, the periods of high server resource usage dictate the total amount of resources which need to be dedicated to the environment on a full-time basis. One of the typical side effects of this approach is long periods of time in which the environment’s system resources are not being utilized or are being underutilized. In a cloud implementation, underutilization means you are paying by the minute for something you are not using!
Allegro Dynamically Scaling ResourcesCoupling the power of Allegro class events with the versatility of a cloud environment, it is possible to dynamically scale your infrastructure in order to have resources available when they are required and turned off when they are no longer needed. Control of the cloud servers is accomplished by integrating your cloud provider’s API into Allegro and then utilizing triggers to turn the cloud servers on and off at the appropriate times. Given the flexibility of Allegro class events, the triggers can range from simple to very complex. A few examples are:
- Turn on additional servers while a particular valuation is running.
- Maintain a ratio of one server on the grid for every 5 users logged into the system.
- Activate additional servers when the current active servers’ system resources become fully utilized and deactivate them when they are no longer being fully utilized.
- Turn on additional servers for a particular period of time each weekday.