Zike Yang丨StreamNative Software Engineer Zike Yang is a Software Engineer at StreamNative, Apache Pulsar Committer, Apache StreamPipes Commiter & PMC Member. His current work primarily focuses on developing Pulsar multi-language clients and improving Pulsar core features. 演讲议题:Apache Pulsar 限流功能在移动云云原生场景下的应用实践 分享时间:8 月 18 日 14:00 - 14:45 议题介绍: 移动云为了实现 Pulsar 在云原生容器化场景下的多租户网络资源隔离,针对 Pulsar 的限流功能、ResourceGroup 以及负载均衡等特性做了大量优化。本次演讲将介绍 Pulsar 的限流实现原理,我们如何优化 ResourceGroup 来实现集群级别的限流,以及在集群级别限流场景下如何优化负载均衡策略。 嘉宾介绍:王嘉凌丨中国移动云能力中心软件开发工程师 中国移动云能力中心软件工程师,Apache Pulsar Contributor,负责移动云消息队列 Pulsar 和 Kafka 的研发。 演讲议题:华为终端云在容器场景中对 Apache Pulsar 的优化实践 分享时间:8 月 18 日 14:45 - 15:30 议题介绍: Apache Pulsar 是一款云原生消息队列,基于其存算分离架构,通常可以在业务低峰期缩容计算层来节省资源。我们在容器化场景下,针对 Apache Pulsar 做了大量优化。如:现在 Pulsar 负载均衡算法依赖于节点过去的负载数据,达到平衡的过程比较缓慢。当开启 HPA,节点在负载均衡的过程中很可能又会触发扩容,而扩容又会引发新的负载均衡。我们要如何优化来让 Pulsar 更加云原生? 嘉宾介绍:林琳丨华为 SDE 专家 华为终端 SDE 专家,Apache Pulsar PMC 成员,拥有近10年中间件与基础架构设计经验,致力于打造稳定可靠的基础设施 演讲议题:Kafka without Zookeeper 分享时间:8 月 18 日 15:45 - 16:30 议题介绍:Currently, Kafka relies on ZooKeeper to store its metadata, ex: brokers info, topics, partitions...etc. KRaft is a new generation of Kafka that runs without Zookeeper. This talk will include:1. Why Kafka needs to develop the new KRaft feature;2. The architectures of the old (with Zookeeper) Kafka and new (without Zookeeper) Kafka;3. Benefit of adopting KRaft;4. How it works internally;5. The monitoring metrics;6. Tools to help troubleshoot issues in KRaft;7. A demo to show what we've achieved so far;8. The roadmap for the Kafka community to move toward KRaft. After this talk, the audience can have better knowledge of what KRaft is, and how it works, and what's the difference with Zookeeper based Kafka, and most importantly, how to monitor it and troubleshoot it. 嘉宾介绍:Luke Chen丨RedHat Senior Software Engineer Luke Chen is a senior software engineer in RedHat working on products to run Apache Kafka on cloud. He is also a committer and PMC member in Apache Kafka. He has been contributed in Apache Kafka for more than 3 years.邓子明丨字节跳动数据开发 国防科技大学,多年数据开发经验,Apache Kafka Committer 演讲议题:Deep Dive the replication protocol in Kafka 分享时间:8 月 18 日 16:30 - 17:15 议题介绍: Being a messaging system, the data durability is very important. The replication ensures automatic failover to other replicas when a server in the cluster fails, so messages remain available in the presence of failures. In Apache Kafka, the replication protocol is not only used to achieve durability, but also to achieve high throughput. In this talk, we'll deep dive how the replication protocol works internally in Kafka. We'll also explain what's the pros and cons of this kind of design. Furthermore, we'll also introduce the other kind of replication protocol in Kafka, which is used for KRaft controllers (i.e. quorum based way). After this talk, the audience can rethink these replication protocols, and maybe some of the ideas can be brought into some other distributed system projects. Hope it will also help audience know more about Apache Kafka. 嘉宾介绍:Luke Chen丨RedHat Senior Software Engineer Luke Chen is a senior software engineer in RedHat working on products to run Apache Kafka on cloud. He is also a committer and PMC member in Apache Kafka. He has been contributed in Apache Kafka for more than 3 years.