Apache Pulsar Community Annual Summit Pulsar Summit Asia 2022 is coming soon

About Pulsar Summit

Pulsar Summit is the annual event of the Apache Pulsar community, which brings together Apache Pulsar project Contributors, Committers, CTO/CIOs, developers, architects, data scientists, and elites from the message and stream computing communities around the world. In this grand event, everyone shares practical experience, exchanges ideas, discusses knowledge about the Pulsar project and community, and interacts.

Pulsar Summit Asia 2022 will be held online on November 19-20, and the conference will be divided into 2 days. The first day is a Chinese forum, and the second day is an English forum, bringing together Asian and global technology experts to share Apache Pulsar practical experience, scenario cases, technology exploration and operation and maintenance stories, and exchange and discuss Pulsar's first-line landing technology dry goods.

At this Pulsar Summit Asia 2022, several technical experts and lecturers from Tencent will share the practice of Apache Pulsar in games, big data, Tencent Cloud, WeChat and other directions.

Get through Apache Pulsar and Envoy to build an efficient game OTO marketing platform practice

Jiang Shuo

Head of Gateway Technology, Tencent Interactive Entertainment GDP Microservice Development Platform

In game marketing, the OTO model (One Time Offer) can well improve user experience and game operation effects. Among them, one of the more common OTO scenarios is to immediately recommend a certain material when a player meets a certain condition, so as to achieve a refined operation effect on users. To achieve this capability cost-effectively, you can use a custom Envoy extension to support the Pulsar protocol. Coupled with Envoy's flexible routing rule configuration and a large number of existing functional plug-ins, various OTO business scenarios can be easily realized through microservices. This speech mainly shares our practical experience in multiple business scenarios, and uses Pulsar, Envoy, and microservices to build an OTO scenario-based marketing platform.

Supporting trillions of data scale, Apache Pulsar's practice in Tencent's big data

Zhang Dawei

Senior Software Engineer, Tencent, Apache Pulsar Committer

At present, Tencent's big data has reached the level of tens of trillions, and the highly reliable and highly consistent scene data has also reached the scale of hundreds of billions. The selection and requirements for MQ are extremely strict. Apache Pulsar relies on cloud-native features and many MQ enterprise-level features in Tencent's big data business has been enriched, and the team has also gained deep precipitation. This sharing will start with multiple scenarios and analyze the landing experience of Apache Pulsar in Tencent's big data business.

Relying on the Geo Replication function to realize the cross-cluster migration function of tenants

Han Mingze

Senior R&D Engineer of Tencent Cloud Middleware, RoP Maintainer

The Geo Replication function is provided in Apache Pulsar to realize inter-cluster data replication. Based on the current cross-cluster data replication capability, we have modified the lookup phase to realize the non-sensing migration function of tenants between clusters. This talk will elaborate on the complete logic of relying on the Geo Replication function to realize cross-cluster migration of tenants. At the same time, here we have also improved the defects of the original subscription replication function. Before migrating the cluster, we added the logic of synchronizing individuallyDeletedMessages to reduce the situation of repeated consumption after the cluster migration.

Practice of Apache Pulsar  in WeChat's large traffic real-time recommendation scenario


Liu Shen

Tencent WeChat Senior R&D Engineer

Apache Pulsar has achieved large-scale production practice in WeChat. This speech mainly introduces the practice of WeChat based on Pulsar in the real-time recommendation scenario of large traffic, including the deployment in K8S in the large traffic scenario, the application of non-persistent topics, load balancing and cache strategy optimization, COS Offloader development and application, etc. . The audience can learn the practical experience of Apache Pulsar in the high-traffic real-time recommendation scenario.

Detailed Apache Pulsar message life cycle

Ran Xiaolong

Senior R&D Engineer of Tencent Cloud Middleware, Apache Pulsar Committer

Topic is abstracted in Apache Pulsar to carry messages sent by users. When a message is sent to Topic, it will be calculated by Broker and then stored in Bookie. This talk will explain in detail how messages are sent to Broker and then processed by Broker and metadata, and finally stored in Bookie. Then it will further explain how Bookie uses the garbage collection mechanism to recycle the data in the Topic, and how the TTL and Retention strategies in the Broker work on the Bookie Client to trigger the garbage collection mechanism.

The principle and practice of Tencent Tianqiong StarRocks natively supporting Pulsar data source


Cheng Guangxu

Technical Director of Tencent Big Data OLAP Platform, member of Apache HBase & InLong PMC

Within Tencent, many users use Pulsar as the main MQ component due to its advantages in performance, usability, and convenience in operation and maintenance; It also attracts more and more users to use it as an OLAP analysis engine; and the powerful combination of the two provides great convenience for Tencent's internal users in data access and analysis.

The main content of the topic includes the detailed explanation of StarRocks real-time data access (routine load), the implementation scheme of Pulsar routine load and the practice of this scheme in Tencent Tianqiong StarRocks.

Click to read  the original text  and sign up for the event now

Registration link: http://hdxu.cn/s8ZHa

Guess you like

Origin blog.csdn.net/TCOS001/article/details/127860389