GCPUG Taipei Meetup #40

GDG Cloud Taipei
Fri, Sep 14, 2018, 7:00 PM (CST)

About this event

1. 於由 Google Taipei 101 Office 的場地限制,Meetup 與會人員需要進行門禁管制(101大樓 + Google Office)
2. 在 Meetup 活動票開放後至截止受理時間,可以自由進行退票不受黑名單機制影響
3. 在 Meetup 活動截止受理後,我們會提交與會人員名單給 Google ,進行 101 大樓貼紙 + Google Office 出入證列印作業。從截止受理時間至活動開始前,沒有出席 + 截止受理時間後退票者,我們全部視為 no show 且列入黑名單中,no show 一次,我們會停權一季

# GCPUG Taipei
大家好,我們是 Google Cloud Platform User Group (GCPUG) 台北分支,我們是一個Google Cloud Platform 相關技術的民間社群,成立的宗旨在分享與交換 Google Cloud Platform 上的一些技術與使用經驗。歡迎對 Google Cloud Platform 有興趣的朋友們可以共襄盛舉。


# 本次活動介紹

- Making Microservices Micro with Istio and Kubernetes
- Knative: A Kubernetes-based platform to manage modern serverless workloads
- From Image Recognition to Object Detection

## 本次活動議程

19:00 - 19:30 報名與集合 (一樓大廳,圓形接待台旁集合)

19:40 - 20:10 Making Microservices Micro with Istio and Kubernetes
> Speaker
RAY TSANG / GCPCloud Developer Advocate
> Abstract
Microservices are here to stay. When applied properly, microservices techniques and culture ultimately help us continuously improve business at a faster pace than traditional architecture. However, microservices architecture itself can be complex to configure. All of a sudden, we are faced with the need for a service discovery server, how do we store service metadata, make decisions on whether to use client side load balancing or server side load balancing, deal with network resiliency, think how do we enforce service policies and audit, trace nested services calls…. The list goes on.

Sure, it’s easy to have a single stack that makes everything work provided there are good microservices support - but what if you have a polyglot environment? How would you make sure all of the stack can address the same concerns in a consistent way? This is where a service mesh comes in.

In this talk, Ray will introduce Istio, an open source service mesh framework created by Google, IBM, and Lyft. We’ll see how the service mesh work, the technology behind it, and how it addresses aforementioned concerns.

20:10 - 20:20 Breaking

20:20 - 20:50 Knative: A Kubernetes-based platform to manage modern serverless workloads
> Speaker
Kyle Bai(白凱仁)​ / 我是 Kyle Bai(白凱仁),目前在 inwinSTACK 擔任軟體開發工程師的研替。自己早期主要開發 iOS 以及 Android Mobile 應用程式,並具備四年專案開發經驗; 而在大四至研所期間專注巨量資料、虛擬化、分散式儲存與容器化等技術研究,如 OpenStack、Hadoop 與 Spark 等,近期主要負責 OpenStack 與 Kubernetes 的整合技術,以及相關應用開發,並利用空閒時間在開源專案做小小貢獻。
> Abstract
Knative 是基於 Kubernetes 平台建構、部署與管理現代 Serverless 工作負載的開源專案,其目標是要幫助雲端供應商與企業平臺營運商替任何雲端環境的開發者、操作者等提供 Serverless 服務體驗。Knative 採用了 Kubernetes 概念來建構函式與應用程式,並以 Istio 實現了叢集內的網路路由,以及進入服務的外部連接,這讓開發者在部署或執行變得更加簡單。本次我將分享 Knative 功能與概念,並實作簡單範例來進行 Demo,讓大家快速體驗。

20:50 - 21:00 Breaking

21:00 - 21:30 From Image Recognition to Object Detection
> Speaker
Tao Chen & Oscar Huang / QNAP Systems Inc.
> Abstract
分享常見的影像處理及影像辨識做法,以及近幾年 AI 界最紅的 Deep Learning,包含深度神經網路(DNN), 卷積神經網路(CNN) 的核心概念,並分享做照片辨識與物件辨識的方法差異,以及如何訓練客製化 YOLO v3 物件偵測模型的過程及踩過的地雷,最後會有個有趣的小 demo

21:30 - End & Networking

Front-end Developer, Back-end Developer, Full-stack Developer, Systems Admin/Ops, Database Admin/Ops, Data Scientist/Analyst, Engineering Manager/Lead, Project Manager, Student, Sales/BD Manager, Startup (co)founder/CEO/CTO, Other