Infrastructure Performance and Applications

IPA team focus is to optimize the speed and efficiency at which LinkedIn delivers content to members through mobile, desktop or tablets. The core telnet is to ensure that every LinkedIn member has a fast and delightful experience. To achieve this, we collect, measure and provide data and insights on the performance metrics for the services within LinkedIn to help improve their stability, scalability and efficiency.

We envision enabling high performant and healthy services, low cost of adoption and raise in developer productivity. To be able to achieve the same, we are working towards building infrastructure to provide performance insights that improve service health and meet our member promise of 24/7 high performance experience at LinkedIn.

IPA comprises of the following groups:

  • Service Performance Insights: The objective of the team is to provide the data and insights related to performance aspects of any service or application within LinkedIn to help optimize its performance and facilitate structured RCA in case of regression. As a part of this objective, we work on the initiatives to provide information related to mapping of topology for service/API interactions and metrics around it. Next, build an application to trace the flow of user requests/calls within the various services for LinkedIn. In addition, our team provides functionality for CPU/Memory profiling on demand for production service at host level. Core systems/tools owned by the team include ODP, callgraph, perf-linter and SPIre

  • Service Infra Experience SI-XP: Service Infrastructure is building the foundation for service development and operations. Developer experience is a core part of the value proposition for Service Infrastructure. Our team focuses on the end-to-end developer experience building, querying and maintaining services