6 Dynamic management of edge application monitoring, measurement, and scheduling

 

This chapter covers

  • Measuring the performance of edge-computing system with key performance indicators
  • Using edge computing to monitor the state of its infrastructure, semantic, and application layers
  • Using the measured and monitored state in scheduling and monitoring edge functions and edge applications

Edge-computing systems can run the gamut between simple embedded device coordination and observation to highly complex distributed collection and processing pipelines. It is crucial to remember—as we’ve discussed in previous chapters—that even in the smallest of use-cases there is an interplay of services and devices the work together to provide the foundation for your edge applications and functions. To ensure your edge-computing system continues to perform optimally, or at all in some cases, it is crucial to keep tabs on how the underlying framework is functioning. To that end, in this chapter we’ll explore how the instrumentation of edge computing can help ensure the continuity of your operations, and help you optimize your own functions as well.

6.1       Measuring performance in edge computing

6.1.1   Key performance indicators

6.1.2   Infrastructure KPI measurements

6.1.3   Application KPI measurements

6.1.4   Associating infrastructure and application KPIs

6.2       Monitoring your system performance

6.2.1   Agent-level monitoring

6.2.2   Application-level monitoring

6.2.3   Region- and global-level monitoring

6.3       Monitoring application events in edge computing

6.3.1   Downloading the CEP plugin

6.3.2   Launching your agent and dashboard

6.3.3   Building your CEP

6.3.4   Uploading the CEP plugin to the agent registry

6.3.5   Adding and configuring plugins in the application builder

6.3.6   Submitting your edge application

6.4       Scheduling functions and applications

6.4.1   Manually scheduling edge components

6.4.2   Scheduling to fit the requirements of the edge function

6.4.3   Scheduling to maximize system performance and utilization

6.5       Summary