An integrated solution for application developers on the Pivotal Cloud Foundary Platform to visualize key performance indicators and troubleshoot issues within logs and traces.
Metrics was the first project I worked on at Pivotal. After a week in the Denver office I was off to San Fransico to meet and work alongside the talented team. They had just spiked on getting application metrics to the browser and were ready for design to an interface to provide insights on performance to developers.
Generative research revealed common patterns in monitoring and troubleshooting application performance. Users were navigating between multiple tools to find the information they needed, increasing the time to find a root cause and resolve the issue. Additionally, the solutions needed were costly and user privileges were difficult and time consuming to manage.
Cody | Application Developer
Needs
• Monitoring and troubleshooting tools
• Support DevOps responsibility principles
• Quick resolution to issues
Pains
• Multiple products required
• User management for access privileges
• Existing tools are very expensive
Prototyping, Feedback, and Iteration
Creating prototypes for user feedback ensured a validated direction and reduced risk. We were able to identify the key performance indicators and determine the proper data visualizations to convey information clearly. We worked to create intuitive interactions, allowing users to quickly narrow their focus on the timeframe of critical incidents. Before long, our team had brought metrics, application events, and logs into one view, providing significantly faster identification of anomalies and correlated incidents. If additional investigation is required, logs can be loaded and filtered for troubleshooting.Ongoing feedback and iteration led us to test and build additional features to further improve the user experience:
• logs search: limit sources, filter, or highlight keyword(s)
• instance vs. aggregate metrics: detail into which instance may be causing an issue
• trace management: visualizes apps and endpoints for a request to expose failures or latency patent #10,073,767
• custom metrics visualizations: user defined measures of work/performance
• alerting: thresholds on metrics that display on charts when breached and notifications sent if desired
• extended data retention: identify trends over time
Faster Issue Resolution
Key metrics, application events, and logs in one view provides a comprehensive view of application health and performance. Users can quickly assess application status and investigate issues within the same interface. Developers using PCF Metrics can reduce their MTTR (mean time to resolution), and in turn, increase efficiency and application uptime.
Affordable, Automatic Access
Once installed, developers can leverage existing permissions within Pivotal Cloud Foundry, to immediately access this integrated solution, effectively eliminating additional administrative effort. The tool comes at no additional cost beyond the footprint to run and store the application generated data. This supports the desired DevOps ideals of developers sharing responsibility for the performance of their applications.
Increased Product Adoption
Downloads of PCF App Metrics continue to grow and many developers rely on this offering to provide insights into the performance of their applications. Several clients elect to display the dashboard for their priority applications on information radiators within their offices for near real time monitoring.
