Design And Implementation Of High Availability Auto Scaling Based On Cloud Platforms

The cloud computing model,which is paid on demand,makes it easier for developers to develop applications that have been running on the best resource configuration server.But in the usual application scenarios,the server may meet the load demand under different circumstances,when the server through the traffic increases,the load increases,we need to add more computing resources to handle the load;when the server through the flow decrease,reducing the load,we need to reduce the computing resources underutilized.Due to the above situations,the cloud platform needs to achieve auto scaling service to ensure the reliable operation of the application,and reduce maintenance costs.Auto scaling is a key service of cloud platform.It can be based on user's defined strategy such as timing or alarm,and make instance expand or contract automatically.In order to guarantee the performance of the system,it can guarantee the performance and reduce the maintenance cost.Based on the understanding and analysis of the complex application scenarios of cloud computing,this paper designs and implements auto scaling based on cloud platforms.In the premise of ensuring the high availability of auto scaling,greatly improve the utilization rate of resources,reduce unnecessary waste of resources or resource surplus,at the same time can be more convenient to deal with the status quo of the server load demand,provides the high quality service,with high throughput,high performance,low latency,high availability.After introducing the cloud computing,virtualization technology,cloud platforms,auto scaling and other related technical and theoretical background,this paper introduces the high availability auto scaling related business requirements and model,achieved the demand analysis.In the process of implementation of auto scaling,firstly,it implements the overall architecture design and database,and then according to the modules of scaling group management,scaling policy management,launch configuration management and trigger management are designed and implemented,which trigger management introduces the timing trigger policy execution and alarm triggering process of policy execution the implementation of the timing task framework,Quartz and monitor the acquisition alarm tool,Ceilometer respectively;in the high availability of automatic extension service,the external load balancing design using Nginx,and uses the consistent hashing algorithm to achieve detection,the list of activities on the cluster node forwarding task downtime and task recovery within the final to achieve high availability service automatic extension.The design and implementation of high availability auto scaling based on cloud platforms,and the function and performance test show that the system can significantly improve the performance and stability of the server cluster,verify the feasibility and effectiveness.