Click to edit Master title style 1 可觀測性 (Observability) 在 Kubernetes Day2 Operation的考量與實踐 E . W. K u o @ i T h o m e K u b e r n e t e s S u m m i t 2 0 2 2 Click to edit Master title style • 緯創資通員工 • 社群的參與者 • 技術的佈道師 • 台灣資料工程協會會員 • Kafka 修煉之道講師 Techlearn 個人技術學習與收集 Click to edit Master title style 3 Agenda Day2 Operation Challenge of Kubernetes Day 2 Operation Tame operational Day2 運營的挑戰 馴服運營 複雜性 可觀測性 實踐與思維 可觀測性 關聯演示 Click to edit Master title style 4 Day2 Operation 定 義 與 說 明 4 Click to edit Master title style 5 Day 2 Operation 的定義 • 一旦“某物”投入運營，Day 2 Operation

3.1 建立 redis-master Pod 及服務 我們可以先定義 Service，然後再定義一個 RC 來建立和控制相對應的 Pod，或者先 定義 RC 來建立 Pod，然後定義與其關聯的 Service，這兩種方式最終的結果都一 樣，這裡我們採用後面這種方式。 首 先 為 redis-master 服 務 建 立 一 個 名 為 redis-master 的 RC 定 義 檔：redis-master- 檔：redis-master- controller.yaml。下面提供該檔的完整內容： apiVersion: v1 kind: ReplicationController metadata: name: redis-master labels: name: redis-master spec: replicas: 1 selector: name: redis-master redis-master template: metadata: labels: name: redis-master 1-42 Kubernetes 入門 1 1.4.9 小結 上述這些元件是 Kubernetes 系統的核心元件，它們共同構成 Kubernetes 系統的框 架和運算模型。透過對它們進行靈活組合，使用者就可快速、方便地對容器叢集進 行配置、建置和管理。

disposal. julia> Threads.nthreads() 4 300CHAPTER 24. MULTI-THREADING 301 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Array{Int64,1}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes. Note realized, while memory associated with globals may be collected when they are reas- signed on the master, no such action is taken on the workers as the bindings continue to be valid. clear! can be used

disposal. julia> Threads.nthreads() 4 300CHAPTER 24. MULTI-THREADING 301 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Array{Int64,1}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes. Note realized, while memory associated with globals may be collected when they are reas- signed on the master, no such action is taken on the workers as the bindings continue to be valid. clear! can be used

disposal. 318CHAPTER 25. MULTI-THREADING 319 julia> Threads.nthreads() 4 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Array{Int64,1}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes.CHAPTER realized, while memory associated with globals may be collected when they are reas- signed on the master, no such action is taken on the workers as the bindings continue to be valid. clear! can be used

disposal. 318CHAPTER 25. MULTI-THREADING 319 julia> Threads.nthreads() 4 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Array{Int64,1}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes.CHAPTER realized, while memory associated with globals may be collected when they are reas- signed on the master, no such action is taken on the workers as the bindings continue to be valid. clear! can be used

disposal. 318CHAPTER 25. MULTI-THREADING 319 julia> Threads.nthreads() 4 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Array{Int64,1}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes.CHAPTER realized, while memory associated with globals may be collected when they are reas- signed on the master, no such action is taken on the workers as the bindings continue to be valid. clear! can be used

(_| | | Version 1.13.0-DEV.639 (2025-05-25) _/ |\__'_|_|_|\__'_| | Commit 925d504ade4 (1 day old master) |__/ | julia> 1 + 2 3 julia> ans 3 To exit the interactive session, type CTRL-D (press the verify there are 4 threads at our disposal. julia> Threads.nthreads() 4 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Vector{Int64}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes.CHAPTER

verify there are 4 threads at our disposal. julia> Threads.nthreads() 4 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Vector{Int64}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes.CHAPTER realized, while memory associated with globals may be collected when they are reas- signed on the master, no such action is taken on the workers as the bindings continue to be valid. clear! can be used

verify there are 4 threads at our disposal. julia> Threads.nthreads() 4 But we are currently on the master thread. To check, we use the function Threads.threadid julia> Threads.threadid() 1 Note If you addprocs(2) 2-element Vector{Int64}: 2 3 Module Distributed must be explicitly loaded on the master process before invoking addprocs. It is automatically made available on the worker processes.CHAPTER realized, while memory associated with globals may be collected when they are reas- signed on the master, no such action is taken on the workers as the bindings continue to be valid. clear! can be used