Scalability

The highly scalable nature of the Internet presents exciting growth prospects. To realize those growth prospects your Internet Infrastructure has to be able to handle the scalability.


OTC uses the Internet Delivery Framework (IDF) to help manage scalability. The IDF helps focus on those components that might be hindering scalability the most while still providing a global view of your scalability.


For more information on scalability call OTC at 734-213-2020, or email sales@onlinetech.net


  • Challenges to Scalability
    • Bandwidth

      • The bandwidth necessary to handle peak visitors to your website can often be many, many times the typical average bandwidth used by your site


      • The challenge is determining how much bandwidth to purchase relative to your peak and your average


      • Adding bandwidth is generally very easy so you don’t have to worry too much about answering "At what point am I willing to turn away customers" for too long


    • Storage

      • Depending on the nature of your site storage could be consumed very, very quickly


      • If you allow end-users to submit content, especially multi-media content, your storage needs can grow exponentially


      • Often adding storage can be a time consuming process that requires downtime and carries some data risk.


    • Servers

      • Application and database servers can handle a certain number of simultaneous users


      • It is heavily dependent on the nature and design of the application and database itself


      • Using multiple servers requires some type of "cluster" or load balancing which your application may need to accomodate



    • Staff

      • Many underestimate the growing staff requirements of a highly scalable environment.


      • As you add servers, bandwidth and storage the number and complexity of devices increases


      • This calls for increased breadth and depth of staff.



  • Tools to enable scalability
    • Framework - A framework such as OTC's Internet Delivery Stack Helps to asssure scalability in all necessary components


    • Load balancing - Spread the load across multiple servers. Place those servers at multiple data centers and gain a level of Disaster Recovery as well.


    • Clusters of servers - With proper design at all layers of the stack you can scale by merely adding physical (or virtual) servers to a "farm" of servers


    • Network storage - When properly sized, network storage can enable incremental and rapid growth


    • Bandwidth - Peak bandwidth can only be achieved with very high speed connections such as gigabit over fiber capacity and/or load distribution.


    • Outsource Staff - As you scale the complexity and variety of technologies you will face will increase exponentially. A common solution is to outsource the infrastructure and the managed services in a way that scales as you scale.