Kebe Liu
DaoCloud, Senior software engineer
Shanghai, China
Actions
Member of Istio Steering Committee, focused on cloud-native and Istio, eBPF and other areas in recent years.
Links
Area of Expertise
Topics
Unified Management, Continuity, Compliance in Multi-Clouds with Service Mesh
In multi-cloud and hybrid cloud architectures, enterprises face challenges like inter-cloud communication, traffic management, application orchestration, data security, and compliance. Service mesh technology offers a unified approach for managing service interactions, enhancing security, and ensuring data compliance.
Istio, a leading service mesh project, is particularly effective in multi-cloud and hybrid cloud environments. It provides seamless network connectivity across various architectures, ensuring reliable and secure communication. Additionally, integrating Istio with Karmada enables efficient application scheduling across these complex environments. Karmada allows for smooth orchestration of workloads across different cloud platforms, enhancing the flexibility and scalability of cloud-native applications.
I aim to share practical insights and experiences, especially from China, to inspire and provide strategic perspectives in navigating these technological landscapes.
Sit back and relax with fault awareness and robust instant recovery for large scale AI workloads
The fault tolerance during train, fine-tuning, and even inferencing is crucial to modern AI workloads when it happens on large scale, with loads of GPU clusters.
For training and fine-tuning tasks, failure of GPUs, storages, any hardware issues often cause the extending the training time to weeks and even months significantly. For inferencing, when massive loads of requests income, if one of the inferencing servers went faulty, we need a policy and scheduler to perform mitigation to transfer the workloads fast and efficiently.
In this talk, We will introduce a series of mechanism we have designed to help Kubernetes clusters and workloads itself to locate, diagnostic the root cause, schedule and perform mitigation when it comes to any of hardware or CUDA API call failures to reduce the overall operating challenges. But the possibilities will not stop here, the fault awareness and mitigation scheduler will help any of the workloads to mitigate during failures.
Istio Open Source Ecosystem Outlook From China（Istio 开源生态展望）
This panel focuses on the development of Istio's open source ecology and technology in China. We will do an in-depth discussion on the hot technologies and open source practices in the Istio field at this stage, such as ebpf, proxyless, custom protocol access, open source products and ecological development, etc.
本次圆桌会议聚焦于Istio在中国的开源生态和技术的发展，我们会对现阶段 Istio 领域的热门技术和开源实践做深入的探讨，例如 ebpf、proxyless、自定义协议接入、开源产品及生态发展等。
Use eBPF instead of iptables to accelerate the Istio dataplane（使用 eBPF 代替 iptables 加速 Istio 数据平面）
Replacing iptables rules with eBPF to allow data being transported directly from inbound sockets to outbound sockets to shorten the datapath between sidecars and services.
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