Runtime Performance Monitoring in Stream Data Applications
|has title::Runtime Performance Monitoring in Streaming Data Applications|
|Master:||project within::Software Engineering|
|Student name:||student name::Andys Sundaypink|
|Supervisor:||Hans van Vliet|
|Second supervisor:||Maurits Kaptein|
|Company:||has company::Science Rockstarts|
Two objectives of providing service over the internet are providing the needed functionality and providing the needed Quality of Service (QoS). QoS is also called as non-functional requirement. The runtime related non-functional requirement is performance oriented. On a cloud-based application perspective, one has to consider the performance during peak time. Performance problems and bottlenecks can cause data losses and application crash. Most likely, customer will lose trust to cloud application whose application has track record to crash. They are most likely reluctant to use the application due to the poor performance and sluggish response time. In order to identify the application performance, monitoring system needs to be in place. An approach to monitor the application performance and what needs to be monitored should be defined. It enables identifying the performance measurement and strategies for dynamically adapt the service performance. The project will investigate and formulate the performance monitoring approach, strategies and challenges in order to identify the performance key and bottleneck of the system. The result will be used as a starting point of next strategy plan whether the system needs to be improved and/or scaled.
In this research work, we would like to be able to identify the bottlenecks which might occur during runtime in the streaming data applications. The Research questions: (1) What are the performance requirements for the streaming data applications? (2) What are monitoring approach to be used in the streaming data applications? What will be the difference with monitoring other type of applications?
The research should covers: (1) Analyse the characteristics, behaviors, and requirements of stream data applications and find out what will be the possible bottleneck based on those attrbutes identified.
(2) Analyse the approach to exploit the application during runtime. It will include the collection of data, understanding each components, make test data.
(3) Find approach to analyse the monitoring data to find the bottlenecks