CLOUD PROPERTIES

CLOUD PROPERTIES

1. Adaptability Property

   • The adaptability property is a statistical number reflecting a Cloud provider’s ability to provide timely and efficient reaction to infrastructure and application changes and events.
     • Example of changes and events include the following: adding new components, removing components, changing the location of components, act of God, an application component failure, etc.
       • The factors that help in assessing the adaptability as virtual and application services include the following:
         • Mean Time to Discover (MTTD) an incident, Mean Time to Invoke (MTTI) an action to remedy the incident Mean Time to Recover (MTTR) from an incident. Trust Tolerance to attack
           • Adaptability Property
             • • Minimizes the mean time to discover a problem (MTTD) • Minimizes the mean time to invoke a solution (MTTI) • Minimizes the mean time to recover a problem (MTTR) • Helps in trust measurement

2. Resilience Property

   • The resilience property is a statistical number reflecting the ability of a system to maintain its features (e.g. serviceability and security) despite a number of sub-system and component failures.
     • High value of resilience requires a design which uses redundancy to eliminate any single point of failure, together with well crafted procedures (e.g. a procedure that defines a disaster recovery process).
       • Resilient design helps in achieving higher availability and reliability, as its design approach focuses on tolerating and surviving the inevitable failures rather than trying to reduce them
         • The factors affecting the value of the resilience property are as follows:
           •  Resilience as virtual and application services  Adaptability property  Tolerance to attack  Architecture  Feedback of availability as virtual and application services and reliability as virtual and application services

3. Scalability property

   • is a statistical figure reflecting the ability of Clouds to enable a virtual infrastructure to scale resources up and down based on demand.
     • For example, obviously, at peak periods the virtual layer should scale resources up, and similarly on off-peak periods the virtual layer should release unneeded resources by scaling down.
       • Scalability at a virtual layer can be of two types:
         • Horizontal scalability. The horizontal scalability is about the amount of instances that would need to be added or removed to a system to satisfy increase or decrease in demand.
           • Vertical scalability. The vertical scalability is about increasing or decreasing the size of instances themselves to maintain increase or decrease in demand.
             • The factors affecting the value of scalability property are as follows:  Scalability as virtual and application services  Tolerance to attack  Failure statistics  Scalability statistics

4. Availability property

   • Availability property is a statistical number reflecting the relative time a service provides its intended functions.
     • High levels of availability property are a result of excellent architecture, which considers well crafted procedures, redundant services, and high service reliability: that is, resilient design.
       • The factors affecting the value of the availability property are as follows:  Resilience property  Availability as virtual and application services  Tolerance to attack  Failure statistics

5. Reliability property

   • is a statistical figure which is related to the success in which a service functions.
     • High end-to-end service reliability implies a service always provides correct results and guarantees no data loss.
       • The factors which affect the Reliability property are as follows:  Resilience property  Reliability as virtual and application services  Tolerance to Attack  Failure Statistics  Integrity incidents

6. business model

   • Users would be interested to understand the values of individual and overall properties of the environment hosting his applications.
     • Knowing the values of individual properties would help Cloud users to assess Cloud providers.
       • A Cloud provider is more interested in providing customers with different service levels based on different group.
         • Each group is associated with hardware and software properties.
           • A Cloud provider might not necessarily maintains a specific and static value for the discussed properties.
             • The option that has higher property values would be more expensive than a group with lower values.
               • A Cloud provider is most likely to provide different options which are associated with appropriate monetary values.
                 • The above would enable Cloud users to decide on the service level they anticipate as agreed in an SLA.