• Created a Kubernetes monitoring developer tool targeting and acquiring node status data to identify and isolate instances of node failure within k8 clusters.
• Leveraged Prometheus and its query language, PromQL in order to generate queries reflecting node health status in a result dependent drop down to fetch targeted status metrics for users involving selecting a specific replica set to monitor and presenting any abnormal spikes in CPU usage above a 5% average range back to the client.
• Implemented React and React Hooks to utilize useState and useEffect for state management in conjunction with React’s component reusability and unidirectional data flow to pass the state of 1 or more detected CPU usage irregularities to child components.
• Detected anomalies are identified based on desired scraping intervals, increasing effectiveness and accuracy over using Kubectl by over 10%.
• Developed the Kubernetes Integration Platform using Node.js and Express, ingesting a stream of k8 cluster CPU utilization, network latency, and pod runtime metrics for SRE teams to monitor their microservices at scales, achieving a 20% reduction in average response time due to efficient routing and management of HTTP requests.
• Employed Docker images to bundle the application along with its dependencies, enhancing workflow and leveraging its isolation capabilities to mitigate potential conflicts within the application when utilized on more than 1 device.
• Developed under tech accelerator OS Labs.
