Haoran Qiu
Research SDE at Microsoft; CS PhD at UIUC
Actions
Haoran Qiu is a Research Software Engineer at Microsoft Azure Systems Research. His research interests are in cloud efficiency, ML systems, and applying ML for cloud systems design and operation. Haoran was a recipient of ML and Systems Rising Star by MLCommons in 2023. Before joining Microsoft, Haoran obtained his PhD degree in Computer Science from UIUC, advised by Prof. Ravishankar Iyer, and obtained his B.Eng. degree in Computer Science from the University of Hong Kong.
Links
The State of Kubernetes Optimization and the Role of AI
Featuring a diverse panel of experts, attendees will hear the latest in Kubernetes optimization. The session will encourage and engage attendees to challenge conventional wisdom and explore innovative approaches to optimization. Participants will leave with actionable knowledge and new perspectives they can apply to their own environments.
Topics include:
- Valuable insights into the current state of AI in optimization, highlighting both its potential and barriers to adoption
- How and when AI can be used for real-time decision-making
- Exploring the intersection of sustainability and optimization, emphasizing the importance of visibility in driving sustainable practices
- The state of multidimensional pod autoscaling and potential to resolve conflicts between horizontal and vertical autoscaling
- How new computing options and tools like Karpenter have the potential to disrupt the bin packing problem
- How cloud-native projects can leverage new tools to track efficiencies
Sustainable Scaling of Kubernetes Workloads with In-Place Pod Resize and Predictive AI
Accurately guessing CPU & memory requirements for workloads is hard! So, it is common for users to over-provision pods which leads to under-utilized clusters, and the need to scale up cluster size to accommodate workloads.
Recently added in-place pod resize feature brings the ability to right-size over-provisioned pods without restarting them. In this talk, Vinay will discuss how cluster autoscaler currently handles pods pending due to insufficient resources, then introduce a change to the autoscaling workflow that right-sizes over-provisioned pods, and show how it can help schedule pending pods more quickly while lowering costs & carbon footprint.
Haoran will talk about the latest research that leverages machine learning and reinforcement learning techniques to achieve multi-dimensional autoscaling, and discuss how this cutting-edge work can help proactively scale workloads to achieve optimal cluster utilization while meeting application SLOs by more precisely provisioning the pods.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top