Auto-sync: 2026-04-19 14:51
This commit is contained in:
34
wiki/concepts/ADOT.md
Normal file
34
wiki/concepts/ADOT.md
Normal file
@@ -0,0 +1,34 @@
|
||||
---
|
||||
title: "ADOT (AWS Distro for OpenTelemetry)"
|
||||
type: concept
|
||||
tags:
|
||||
- OpenTelemetry
|
||||
- AWS
|
||||
- ADOT
|
||||
date: 2024-04-02
|
||||
---
|
||||
|
||||
## Definition
|
||||
ADOT(AWS Distro for OpenTelemetry)是 AWS 提供的 OpenTelemetry 发行版,作为统一代理用于收集 Traces、Metrics、Logs,并自动检测应用语言创建预配置的 OpenTelemetry Collector。
|
||||
|
||||
## Key Features
|
||||
- **统一代理**:单一 Agent 收集所有类型遥测数据
|
||||
- **自动检测**:自动识别应用编程语言并配置 instrumentation
|
||||
- **AWS 集成**:原生支持 CloudWatch、X-Ray、OpenSearch、Prometheus 等 AWS 服务
|
||||
- **Operator 支持**:通过 Kubernetes Operator 自动管理 Collector 部署
|
||||
|
||||
## Capabilities
|
||||
- 自动 instrumentation(自动埋点)
|
||||
- 自定义属性添加(如租户 ID)
|
||||
- 日志支持(Fluent Bit 集成)
|
||||
- 无服务器指标抓取(Amazon Managed Prometheus)
|
||||
|
||||
## Related Components
|
||||
- [[OpenTelemetry]]:上游开源项目
|
||||
- [[OpenTelemetry-Collector]]:ADOT 基于 Collector 构建
|
||||
- [[EKS]]:主要部署平台
|
||||
- [[Amazon-OpenSearch-Service]]:数据存储后端
|
||||
|
||||
## References
|
||||
- AWS re:Invent 演示:EKS 上使用 Fluent Bit 收集日志,转发至 OpenTelemetry Collector 端点(端口 55681)
|
||||
- 支持 11 种语言 SDK
|
||||
18
wiki/concepts/BYOD.md
Normal file
18
wiki/concepts/BYOD.md
Normal file
@@ -0,0 +1,18 @@
|
||||
---
|
||||
title: "BYOD"
|
||||
type: concept
|
||||
tags: []
|
||||
---
|
||||
|
||||
## Definition
|
||||
Bring Your Own Device,自带设备。允许员工使用个人设备(笔记本、手机、平板)访问企业资源的工作模式。
|
||||
|
||||
## Security Considerations
|
||||
- 终端数据保护
|
||||
- 企业数据与个人数据隔离
|
||||
- 远程擦除能力
|
||||
- 多因素认证
|
||||
|
||||
## Related Concepts
|
||||
- [[VDI]]
|
||||
- [[SAML]]
|
||||
24
wiki/concepts/BottleRocket.md
Normal file
24
wiki/concepts/BottleRocket.md
Normal file
@@ -0,0 +1,24 @@
|
||||
---
|
||||
title: "BottleRocket"
|
||||
description: EKS Auto Mode 使用的 Linux 操作系统,专为容器工作负载优化
|
||||
type: concept
|
||||
tags:
|
||||
- AWS
|
||||
- EKS
|
||||
- Operating-System
|
||||
---
|
||||
|
||||
## Definition
|
||||
BottleRocket 是 AWS 开发的 Linux 操作系统,专为容器工作负载设计,被 EKS Auto Mode 采用作为默认操作系统。它支持自动化补丁管理和安全更新。
|
||||
|
||||
## Features
|
||||
- 专为容器优化的轻量级 OS
|
||||
- 自动化补丁管理和安全更新
|
||||
- 与 EKS Auto Mode 深度集成
|
||||
|
||||
## Relationship
|
||||
- 由 [[Carpenter]] 管理
|
||||
- 运行在 [[EKS-Auto-Mode]] 集群节点上
|
||||
|
||||
## References
|
||||
- [[public-cloud-learning-sessions-eks-optimization-part-3-of-3-introduction-to-eks-auto-mode]]
|
||||
37
wiki/concepts/Bottlerocket-OS.md
Normal file
37
wiki/concepts/Bottlerocket-OS.md
Normal file
@@ -0,0 +1,37 @@
|
||||
---
|
||||
title: "Bottlerocket OS"
|
||||
type: concept
|
||||
tags: [AWS, Container, Operating-System, Security]
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
Bottlerocket OS 是专为托管容器工作负载而设计的最小化 Linux 操作系统,与通用操作系统不同,仅包含必要组件。
|
||||
|
||||
## Key Characteristics
|
||||
- **最小化设计**:无包管理器、无默认 shell 解释器、无默认 SSH 访问,仅包含必要的内核组件
|
||||
- **变体机制**:通过变体满足特定工作负载需求(如 GPU 支持)
|
||||
- **安全更新**:原地更新和节点替换两种方式,使用 dm-verity 验证根文件系统完整性
|
||||
- **不可变根文件系统**:根文件系统默认不可变,/etc 是临时文件系统
|
||||
- **SELinux**:默认强制启用
|
||||
- **CIS Benchmark**:提供专门的硬化基准
|
||||
|
||||
## Variants
|
||||
Bottlerocket 变体是平台、处理器架构和必要二进制组件的组合:
|
||||
- 基础变体
|
||||
- GPU 变体(支持 NVIDIA 驱动)
|
||||
- 与 EKS、Carpenter 集成的优化变体
|
||||
|
||||
## Use Cases
|
||||
- EKS 节点操作系统
|
||||
- 自托管 Kubernetes 集群
|
||||
- 容器化工作负载生产环境
|
||||
|
||||
## Related Concepts
|
||||
- [[dm-verity]] — 根文件系统完整性验证
|
||||
- [[CIS-Benchmarks]] — 安全配置基准
|
||||
- [[EKS]] — 支持的 Kubernetes 服务
|
||||
|
||||
## Related Entities
|
||||
- [[Bottlerocket]] — 维护的开源项目
|
||||
- [[AWS]] — 核心维护者和赞助商
|
||||
25
wiki/concepts/Carpenter.md
Normal file
25
wiki/concepts/Carpenter.md
Normal file
@@ -0,0 +1,25 @@
|
||||
---
|
||||
title: "Carpenter"
|
||||
description: EKS Auto Mode 的计算控制器,负责基础设施管理和节点生命周期
|
||||
type: concept
|
||||
tags:
|
||||
- AWS
|
||||
- EKS
|
||||
- Controller
|
||||
---
|
||||
|
||||
## Definition
|
||||
Carpenter 是 EKS Auto Mode 的计算控制器,运行在作为核心能力的 Pod 中,负责管理节点生命周期、基础设施配置和 AMI 滚动更新。
|
||||
|
||||
## Functionality
|
||||
- 管理节点池配置和实例类型选择
|
||||
- 响应控制平面版本变化,自动拉取新 AMI
|
||||
- 通过滚动升级跨集群部署新 AMI
|
||||
- 支持动态实例确定和自动整合
|
||||
|
||||
## Relationship
|
||||
- 与 [[EKS-Auto-Mode]] 紧密集成,是其核心能力之一
|
||||
- 管理 [[BottleRocket]] 操作系统实例
|
||||
|
||||
## References
|
||||
- [[public-cloud-learning-sessions-eks-optimization-part-3-of-3-introduction-to-eks-auto-mode]]
|
||||
29
wiki/concepts/Cluster-Autoscaler.md
Normal file
29
wiki/concepts/Cluster-Autoscaler.md
Normal file
@@ -0,0 +1,29 @@
|
||||
---
|
||||
title: "Cluster Autoscaler"
|
||||
type: concept
|
||||
tags: [Kubernetes, Auto-Scaling, GKE, EKS]
|
||||
sources: []
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
Cluster Autoscaler 是 Kubernetes 社区的自动扩缩容组件,与 Karpenter 功能类似但实现方式不同。
|
||||
|
||||
## Key Differences from Karpenter
|
||||
- 不直接与 EC2 Fleet API 通信,依赖 Kubernetes scheduler
|
||||
- 无原生 Spot 中断处理,需要额外组件(如 Node Termination Handler)
|
||||
- 节点管理通过 Node Groups 而非原生资源
|
||||
- 不支持工作负载放置的细粒度控制
|
||||
|
||||
## Use Cases
|
||||
- GKE:Google Kubernetes Engine 原生支持
|
||||
- AKS:Azure Kubernetes Service 支持
|
||||
- EKS:需要额外配置
|
||||
|
||||
## Related Concepts
|
||||
- [[Karpenter]]:Cluster Autoscaler 的替代方案
|
||||
- [[Node-Pools]]:类似 Karpenter 的节点管理概念
|
||||
|
||||
## Aliases
|
||||
- Kubernetes Cluster Autoscaler
|
||||
- CA
|
||||
27
wiki/concepts/Comparative-Mode.md
Normal file
27
wiki/concepts/Comparative-Mode.md
Normal file
@@ -0,0 +1,27 @@
|
||||
---
|
||||
title: "Comparative Mode"
|
||||
type: concept
|
||||
tags: [comparison, research, mode]
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
|
||||
Comparative Mode 是 Last30Days skill 的一种研究模式,通过 "X vs Y" 格式生成并排对比研究,帮助用户快速对比两个主题的热门内容、用户痛点和市场观点。
|
||||
|
||||
## Usage
|
||||
|
||||
```bash
|
||||
python3 ~/.openclaw/skills/last30days-official/scripts/last30days.py "cursor vs windsurf"
|
||||
```
|
||||
|
||||
## Output Structure
|
||||
|
||||
- 双方热门内容并排显示
|
||||
- 关键模式对比
|
||||
- 推荐行动并列
|
||||
|
||||
## Related Concepts
|
||||
|
||||
- [[Last-30-Days-Skill]]
|
||||
- [[Market-Research]]
|
||||
30
wiki/concepts/Consolidation-Policies.md
Normal file
30
wiki/concepts/Consolidation-Policies.md
Normal file
@@ -0,0 +1,30 @@
|
||||
---
|
||||
title: "Consolidation Policies"
|
||||
type: concept
|
||||
tags: [Kubernetes, Karpenter, Cost-Optimization]
|
||||
sources: []
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
Consolidation Policies 是 Karpenter 的成本优化策略,控制节点的整合行为以减少空闲资源。
|
||||
|
||||
## Configuration Options
|
||||
- **Budget-based**:基于预算的整合限制
|
||||
- **Time-based**:基于时间的整合控制(如业务高峰时段禁止整合)
|
||||
- **Percentage-based**:限制每次中断的实例百分比
|
||||
|
||||
## Purpose
|
||||
- 在保证性能的前提下最小化成本
|
||||
- 避免在业务高峰时段进行节点整合
|
||||
- 控制变更节奏,减少对工作负载的影响
|
||||
|
||||
## Related Concepts
|
||||
- [[Karpenter]]:使用 Consolidation Policies
|
||||
- [[Node-Pools]]:配置整合策略的组件
|
||||
- [[Cost-Optimization]]:成本优化
|
||||
|
||||
## Aliases
|
||||
- Consolidation Policies
|
||||
- Consolidation
|
||||
- Node Consolidation
|
||||
22
wiki/concepts/Dashboard-as-Code.md
Normal file
22
wiki/concepts/Dashboard-as-Code.md
Normal file
@@ -0,0 +1,22 @@
|
||||
---
|
||||
title: "Dashboard-as-Code"
|
||||
type: concept
|
||||
tags: [Grafana, IaC, monitoring, automation]
|
||||
sources: [ctp-topic-60-monitor-aws-using-hyperscale-observability-with-grafana]
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Summary
|
||||
通过代码管理 Grafana 仪表板的实践,使用 Terraform 模块实现自动化 provisioning。
|
||||
|
||||
## Definition
|
||||
用代码(通常是 Terraform HCL)定义和管理 Grafana 组织、用户、文件夹、IAM 角色和仪表板的实践。
|
||||
|
||||
## Key Attributes
|
||||
- **类型**:基础设施即代码(IaC)
|
||||
- **工具**:Terraform
|
||||
- **用途**:自动化监控资源部署
|
||||
|
||||
## Connections
|
||||
- [[Grafana]] ← 管理工具 ← [[Dashboard-as-Code]]
|
||||
- [[Terraform]] ← 使用模块 ← [[Dashboard-as-Code]]
|
||||
60
wiki/concepts/EKS-可靠性.md
Normal file
60
wiki/concepts/EKS-可靠性.md
Normal file
@@ -0,0 +1,60 @@
|
||||
---
|
||||
title: "EKS 可靠性"
|
||||
type: concept
|
||||
tags: [AWS, EKS, Kubernetes, Reliability]
|
||||
---
|
||||
|
||||
## Description
|
||||
EKS 可靠性是指在 Amazon EKS 集群中实现高可用性和弹性的实践,确保系统在故障发生时仍提供可预测行为。EKS 可靠性涵盖三个层面:应用可靠性、控制平面可靠性、数据平面可靠性。
|
||||
|
||||
## Three Layers of EKS Reliability
|
||||
|
||||
### 1. 应用可靠性(Application Reliability)
|
||||
- 避免单点 Pod,使用 [[Pod 反亲和性]] 或 [[拓扑分布约束]]
|
||||
- [[HPA]](Horizontal Pod Autoscaler)根据 CPU/内存自动扩展
|
||||
- [[VPA]](Vertical Pod Autoscaler)自动调整资源请求
|
||||
- [[探针]](Liveness、Readiness、Startup)监控 Pod 健康
|
||||
- [[Pod 中断预算]] 确保维护期间最低服务水平
|
||||
- 部署策略:Rolling、Blue-Green、Canary
|
||||
|
||||
### 2. 控制平面可靠性(Control Plane Reliability)
|
||||
- 监控控制平面指标(API Server 请求、etcd 状态)
|
||||
- 安全认证配置
|
||||
- 精心配置和测试的 Admission Webhooks
|
||||
- 集群升级:控制平面和数据平面分阶段升级
|
||||
- EKS 平台版本自动透明升级
|
||||
- minor 版本 14 个月支持周期后自动升级
|
||||
|
||||
### 3. 数据平面可靠性(Data Plane Reliability)
|
||||
- 节点问题检测器(Node Problem Detector)
|
||||
- 系统资源预留
|
||||
- 实施 QoS(Quality of Service)资源配额
|
||||
- 资源限制范围(LimitRanges)
|
||||
- Pod 优先级和抢占
|
||||
|
||||
## Shared Responsibility Model
|
||||
根据 AWS 共享责任模型:
|
||||
- **AWS 负责**:控制平面组件(etcd、API Server、Scheduler、Controller Manager)
|
||||
- **客户负责**:Worker Node、操作系统、应用配置
|
||||
- **Fargate 模式**:AWS 负责节点管理和补丁升级
|
||||
|
||||
## Related Entities
|
||||
- [[Surav Paul]]:演讲人
|
||||
- [[EKS]]
|
||||
- [[AWS]]
|
||||
- [[Fargate]]
|
||||
- [[Kubernetes]]
|
||||
|
||||
## Related Concepts
|
||||
- [[Pod 反亲和性]]
|
||||
- [[拓扑分布约束]]
|
||||
- [[HPA]]
|
||||
- [[VPA]]
|
||||
- [[探针]]
|
||||
- [[Pod 中断预算]]
|
||||
- [[共享责任模型]]
|
||||
|
||||
## Connections
|
||||
- [[EKS 可靠性]] ← 包含 [[Pod 反亲和性]]、[[拓扑分布约束]]、[[HPA]]、[[VPA]]、[[探针]]、[[Pod 中断预算]]
|
||||
- [[EKS]] ← 提供 [[EKS 可靠性]]
|
||||
- [[Surav Paul]] ← 阐述 [[EKS 可靠性]]
|
||||
25
wiki/concepts/ELK-Stack.md
Normal file
25
wiki/concepts/ELK-Stack.md
Normal file
@@ -0,0 +1,25 @@
|
||||
---
|
||||
title: "ELK Stack"
|
||||
type: concept
|
||||
tags: [Log-Analytics, Open-Source, Elasticsearch, Logstash, Kibana]
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Definition
|
||||
ELK Stack 是开源日志分析技术栈,由 Elasticsearch、Logstash 和 Kibana 三个组件组成,用于日志采集、存储、搜索和可视化。
|
||||
|
||||
## Components
|
||||
- **Elasticsearch**:分布式搜索引擎和存储引擎,用于存储和搜索日志数据
|
||||
- **Logstash**:日志处理管道,负责日志的聚合、转换和 enrichment
|
||||
- **Kibana**:Web 前端,用于日志数据的可视化、分析和查询
|
||||
|
||||
## Usage
|
||||
ELK Stack 是云环境日志分析的标准开源方案,通过 BEATS 采集日志,Logstash 处理,Elasticsearch 存储,Kibana 可视化。
|
||||
|
||||
## Connections
|
||||
- [[ELK Stack]] ← uses ← [[BEATS]]
|
||||
- [[ELK Stack]] ← uses ← [[Logstash]]
|
||||
- [[ELK Stack]] ← uses ← [[Elasticsearch]]
|
||||
- [[ELK Stack]] ← uses ← [[Kibana]]
|
||||
- [[OpenSearch]] ← extends ← [[ELK Stack]]
|
||||
- [[Log Analytics]] ← implements ← [[ELK Stack]]
|
||||
29
wiki/concepts/Fargate.md
Normal file
29
wiki/concepts/Fargate.md
Normal file
@@ -0,0 +1,29 @@
|
||||
---
|
||||
title: "Fargate"
|
||||
type: concept
|
||||
tags: [AWS, Serverless, Container]
|
||||
---
|
||||
|
||||
## Description
|
||||
Fargate 是 AWS 提供的无服务器容器运行环境,让用户无需管理底层服务器即可运行容器。在 Fargate 模式下,AWS 负责管理节点、操作系统和补丁升级。
|
||||
|
||||
## Features
|
||||
- 无需预置或管理服务器
|
||||
- 按实际使用的计算资源付费
|
||||
- AWS 自动处理节点管理和安全补丁
|
||||
- 与 EKS 和 ECS 集成
|
||||
|
||||
## Use Case
|
||||
在 EKS 中使用 Fargate 运行无状态工作负载,让 AWS 管理底层基础设施,专注于应用逻辑。
|
||||
|
||||
## Related Entities
|
||||
- [[AWS]]
|
||||
- [[EKS]]
|
||||
- [[ECS]]
|
||||
|
||||
## Related Concepts
|
||||
- [[EKS 可靠性]]
|
||||
- [[Serverless-Computing]]
|
||||
|
||||
## Connections
|
||||
- [[Fargate]] ← 提供 [[EKS 可靠性]]
|
||||
21
wiki/concepts/HPA.md
Normal file
21
wiki/concepts/HPA.md
Normal file
@@ -0,0 +1,21 @@
|
||||
---
|
||||
title: "HPA"
|
||||
type: concept
|
||||
tags: [Kubernetes, EKS, Auto-Scaling]
|
||||
---
|
||||
|
||||
## Description
|
||||
HPA(Horizontal Pod Autoscaler)是 Kubernetes 的水平 Pod 自动扩缩容功能,根据 CPU、内存或自定义指标自动调整 Pod 副本数。
|
||||
|
||||
## Related Concepts
|
||||
- [[VPA]]
|
||||
- [[EKS 可靠性]]
|
||||
- [[KEDA]]
|
||||
- [[Auto-scaling]]
|
||||
|
||||
## Related Entities
|
||||
- [[EKS]]
|
||||
- [[Kubernetes]]
|
||||
|
||||
## Connections
|
||||
- [[HPA]] ← 实现 [[EKS 可靠性]]
|
||||
31
wiki/concepts/IAM-用户.md
Normal file
31
wiki/concepts/IAM-用户.md
Normal file
@@ -0,0 +1,31 @@
|
||||
---
|
||||
title: "IAM 用户"
|
||||
type: concept
|
||||
tags: [AWS, IAM, Identity]
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
IAM 用户是 AWS IAM 中的持久化身份,代表使用 AWS 资源的人员或应用程序。
|
||||
|
||||
## Characteristics
|
||||
- 长期凭证(Access Key + Secret Key)
|
||||
- 可直接附加策略
|
||||
- 适用于服务账号而非人员
|
||||
|
||||
## Use Cases
|
||||
- 服务间通信的凭证
|
||||
- CI/CD 管道的访问凭证
|
||||
|
||||
## Best Practice
|
||||
- 优先使用联合访问替代 IAM 用户进行人员认证
|
||||
- IAM 用户仅用于非人员实体(服务账号)
|
||||
|
||||
## Related Concepts
|
||||
- [[IAM-角色]]: 临时凭证,适用于人员和服务的短期访问
|
||||
- [[IAM-策略]]: 定义 IAM 用户可执行的操作
|
||||
- [[联合访问]]: 优先使用的人员访问方式
|
||||
|
||||
## Connections
|
||||
- [[IAM-用户]] ← uses ← [[IAM-策略]]
|
||||
- [[IAM-用户]] ← alternative_to ← [[联合访问]]
|
||||
45
wiki/concepts/IAM-策略.md
Normal file
45
wiki/concepts/IAM-策略.md
Normal file
@@ -0,0 +1,45 @@
|
||||
---
|
||||
title: "IAM 策略"
|
||||
type: concept
|
||||
tags: [AWS, IAM, Security, Policy]
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
IAM 策略是定义 AWS 权限的 JSON 文档,指定允许或拒绝的操作和资源。
|
||||
|
||||
## Core Concept
|
||||
> "We only want to allow the access that is strictly required."
|
||||
|
||||
最小权限原则是 IAM 策略设计的核心指导原则。
|
||||
|
||||
## Types
|
||||
- **AWS 托管策略**:AWS 预定义的策略,可重用
|
||||
- **客户托管策略**:用户创建和维护的策略,可重用
|
||||
- **内联策略**:直接嵌入 IAM 角色或用户,不可重用
|
||||
|
||||
## Best Practices
|
||||
- 使用内联策略进行角色特定的权限
|
||||
- 使用托管策略进行跨角色可重用的权限
|
||||
- 策略应细粒度,限制访问特定资源而非广泛开放
|
||||
|
||||
## JSON Structure
|
||||
```json
|
||||
{
|
||||
"Version": "2012-10-17",
|
||||
"Statement": [{
|
||||
"Effect": "Allow|Deny",
|
||||
"Action": ["s3:GetObject", "s3:ListBucket"],
|
||||
"Resource": "arn:aws:s3:::bucket-name/*"
|
||||
}]
|
||||
}
|
||||
```
|
||||
|
||||
## Related Concepts
|
||||
- [[IAM-角色]]: 策略附加的目标
|
||||
- [[内联策略]]: 绑定到特定角色
|
||||
- [[托管策略]]: 可跨角色重用
|
||||
- [[最小权限原则]]: 策略设计原则
|
||||
|
||||
## Connections
|
||||
- [[IAM-策略]] ← attached_to ← [[IAM-角色]]
|
||||
40
wiki/concepts/IAM-角色.md
Normal file
40
wiki/concepts/IAM-角色.md
Normal file
@@ -0,0 +1,40 @@
|
||||
---
|
||||
title: "IAM 角色"
|
||||
type: concept
|
||||
tags: [AWS, IAM, Identity]
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
IAM 角色是 AWS IAM 中的身份,可以被其他实体 assum 以获取临时安全凭证。
|
||||
|
||||
## Core Concept
|
||||
> "Roles don't enable actions; they tie together who can do something and what they can do."
|
||||
|
||||
角色本身不执行操作,而是将"谁可以做什么"关联在一起。
|
||||
|
||||
## Characteristics
|
||||
- 临时凭证(通过 AssumeRole API 获取)
|
||||
- 可被服务(AWS Service Role)或用户 assum
|
||||
- 包含信任策略和权限策略
|
||||
- 无持久化登录凭证
|
||||
|
||||
## Use Cases
|
||||
- 授予服务访问 AWS 资源的权限(如 EC2 实例角色)
|
||||
- 授予用户跨账号访问权限
|
||||
- 联合访问映射(AD 组 → IAM 角色 → 权限)
|
||||
|
||||
## Types
|
||||
- **服务角色**:供 AWS 服务使用
|
||||
- **跨账号角色**:跨 AWS 账号授权
|
||||
- **联合角色**:外部身份提供商映射的角色
|
||||
|
||||
## Related Concepts
|
||||
- [[IAM-策略]]: 定义角色可执行的操作
|
||||
- [[信任策略]]: 定义谁可以 assum 角色
|
||||
- [[联合访问]]: AD 组映射到 IAM 角色的工作流
|
||||
- [[最小权限原则]]: 策略设计原则
|
||||
|
||||
## Connections
|
||||
- [[IAM-角色]] ← contains ← [[IAM-策略]]
|
||||
- [[IAM-角色]] ← defined_by ← [[信任策略]]
|
||||
30
wiki/concepts/IPv6-Networking.md
Normal file
30
wiki/concepts/IPv6-Networking.md
Normal file
@@ -0,0 +1,30 @@
|
||||
---
|
||||
title: "IPv6 Networking"
|
||||
type: concept
|
||||
tags: [Networking, VPC, AWS, IP-Address]
|
||||
---
|
||||
|
||||
## Description
|
||||
IPv6 Networking(IPv6 网络)是解决 VPC IP 地址耗尽的方案。通过双栈 VPC,节点支持 IPv4+IPv6 双协议栈,Pod 仅使用 IPv6 地址。IPv6 Pod 与 IPv4 目标交互需要在两层进行 NAT 转换。
|
||||
|
||||
## AWS Implementation
|
||||
- 使用双栈 VPC
|
||||
- 节点支持双栈 IP 地址
|
||||
- Pod 仅分配 IPv6 地址
|
||||
- 通过 NAT 解决 IPv6 与 IPv4 互操作
|
||||
|
||||
## Use Cases
|
||||
- 解决大规模 Kubernetes 集群的 IP 耗尽问题
|
||||
- 支持更多 Pod 和服务
|
||||
- VPC 网络扩展
|
||||
|
||||
## Related Concepts
|
||||
- [[VPC]]
|
||||
- [[EKS]]
|
||||
|
||||
## Related Entities
|
||||
- [[AWS]]
|
||||
|
||||
## Connections
|
||||
- [[IPv6 Networking]] ← resolves [[IP Exhaustion]]
|
||||
- [[IPv6 Networking]] ← extends [[VPC]]
|
||||
20
wiki/concepts/Identity-Governance.md
Normal file
20
wiki/concepts/Identity-Governance.md
Normal file
@@ -0,0 +1,20 @@
|
||||
---
|
||||
title: "Identity-Governance"
|
||||
type: concept
|
||||
tags: []
|
||||
---
|
||||
|
||||
## Definition
|
||||
身份治理(Identity Governance)是一个用于高效管理数字身份、降低风险并保持合规性的框架。它回答三个核心问题:谁当前有权访问我们的系统?谁应该有权访问?访问是如何进行的?
|
||||
|
||||
## Components
|
||||
- 身份管理(Identity Management)
|
||||
- 访问管理(Access Management)
|
||||
- 身份审计(Identity Auditing)
|
||||
|
||||
## Use Cases
|
||||
- 管理内部员工访问权限
|
||||
- 管理外部用户(包括合同工)访问权限
|
||||
- 支持时间限制的临时访问
|
||||
- 通过工作流自动化访问审批和撤销
|
||||
- 监控和审计访问行为
|
||||
30
wiki/concepts/KEDA.md
Normal file
30
wiki/concepts/KEDA.md
Normal file
@@ -0,0 +1,30 @@
|
||||
---
|
||||
title: "KEDA"
|
||||
type: concept
|
||||
tags: [Kubernetes, EKS, Auto-Scaling, Event-Driven]
|
||||
---
|
||||
|
||||
## Description
|
||||
KEDA(Kubernetes Event-driven Autoscaling)是基于外部事件的 Kubernetes 工作负载扩缩容框架。它使用 ScaledObject CRD 定义扩缩容规则,支持从零副本启动,可发布指标供 HPA 使用。
|
||||
|
||||
## Mechanism
|
||||
- 使用 ScaledObject 自定义资源定义
|
||||
- 支持多种外部事件源(消息队列、HTTP API、Azure Blob 等)
|
||||
- 可与 HPA 集成提供指标
|
||||
- 支持从零副本扩展
|
||||
|
||||
## Use Cases
|
||||
- 基于消息队列深度自动扩缩容
|
||||
- 基于 HTTP 请求速率扩缩容
|
||||
- 事件驱动的工作负载
|
||||
|
||||
## Related Concepts
|
||||
- [[HPA]]
|
||||
- [[Auto-scaling]]
|
||||
|
||||
## Related Entities
|
||||
- [[EKS]]
|
||||
|
||||
## Connections
|
||||
- [[KEDA]] ← integrates_with [[HPA]]
|
||||
- [[KEDA]] ← provides_metrics_for [[Auto-scaling]]
|
||||
23
wiki/concepts/Kibana.md
Normal file
23
wiki/concepts/Kibana.md
Normal file
@@ -0,0 +1,23 @@
|
||||
---
|
||||
title: "Kibana"
|
||||
type: concept
|
||||
tags: [Log-Analytics, Visualization, Elasticsearch]
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Definition
|
||||
Kibana 是 ELK Stack 的 Web 前端和可视化界面,用于日志数据的搜索、分析和可视化展示。
|
||||
|
||||
## Description
|
||||
Kibana 提供交互式仪表板、时间序列可视化、图形图表等功能,用户可以通过查询语法搜索日志、创建可视化图表、构建仪表板。End users can view logs via Kibana, connecting from a specified network.
|
||||
|
||||
## Usage
|
||||
- 日志搜索和查询
|
||||
- 可视化仪表板构建
|
||||
- 告警规则配置
|
||||
- 权限管理(配合 RBAC)
|
||||
|
||||
## Connections
|
||||
- [[Kibana]] ← connects_to ← [[Elasticsearch]]
|
||||
- [[ELK Stack]] ← depends_on ← [[Kibana]]
|
||||
- [[Log Analytics]] ← uses ← [[Kibana]]
|
||||
41
wiki/concepts/Log-Analytics.md
Normal file
41
wiki/concepts/Log-Analytics.md
Normal file
@@ -0,0 +1,41 @@
|
||||
---
|
||||
title: "Log Analytics"
|
||||
type: concept
|
||||
tags: [Log-Analytics, Observability, DevOps]
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Definition
|
||||
Log Analytics(日志分析)是云运维可观测性的核心组件,负责日志数据的采集、存储、搜索和可视化,帮助运维团队监控系统健康、排查故障和安全审计。
|
||||
|
||||
## Architecture
|
||||
典型日志分析架构包含:
|
||||
1. **采集层**:BEATS(Filebeat、Metricbeat、Heartbeat 等)从应用采集日志
|
||||
2. **处理层**:Logstash 聚合和转换日志数据
|
||||
3. **存储层**:Elasticsearch 或 OpenSearch 存储和索引日志
|
||||
4. **可视化层**:Kibana 提供查询和可视化界面
|
||||
5. **可选缓冲**:Redis 防止 Logstash 过载
|
||||
|
||||
## Security Measures
|
||||
- 静态加密:加密节点 + NVMe 设备硬件级加密
|
||||
- 传输加密:TLS 1.2
|
||||
- VPC 间私有流量,不经过公网
|
||||
- 基于索引的访问控制 + RBAC
|
||||
|
||||
## Regional Deployment
|
||||
出于 GDPR 合规要求,日志农场按区域 split(Oregon 美国、Europe 欧洲)。
|
||||
|
||||
## Solutions Comparison
|
||||
| 方案 | 成本(单农场/14天/100GB日) | SLA | 特点 |
|
||||
|------|---------------------------|-----|------|
|
||||
| Logz.io | ~$4,000/月 | 99.8% | 托管 ELK,试用期 |
|
||||
| AWS OpenSearch | ~$1,500/月 | 99.9% | 托管,自动快照 |
|
||||
| 自托管 ELK | 最低 | 自定义 | 维护量大 |
|
||||
| Microfocus OBA | 较高 | 成熟 | 商业选项,自动化集群 |
|
||||
|
||||
## Connections
|
||||
- [[Log Analytics]] ← implements ← [[Observability-Engineering]]
|
||||
- [[Log Analytics]] ← uses ← [[ELK Stack]]
|
||||
- [[Log Analytics]] ← uses ← [[OpenSearch]]
|
||||
- [[ELK Stack]] ← provides ← [[Log Analytics]]
|
||||
- [[OpenSearch]] ← provides ← [[Log Analytics]]
|
||||
18
wiki/concepts/Logstash.md
Normal file
18
wiki/concepts/Logstash.md
Normal file
@@ -0,0 +1,18 @@
|
||||
---
|
||||
title: "Logstash"
|
||||
type: concept
|
||||
tags: [Log-Analytics, Data-Processing, ETL]
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Definition
|
||||
Logstash 是 ELK Stack 中的日志处理管道组件,负责接收、转换和 enrichment 日志数据,然后发送到 Elasticsearch 存储。
|
||||
|
||||
## Description
|
||||
Logstash 支持多种输入源(文件、网络、Beats 等),通过过滤器对日志进行解析、转换、添加字段等处理,然后输出到目标存储。可选使用 Redis 作为消息队列缓冲,防止 Logstash 过载。
|
||||
|
||||
## Connections
|
||||
- [[Logstash]] ← receives_from ← [[BEATS]]
|
||||
- [[Logstash]] ← sends_to ← [[Elasticsearch]]
|
||||
- [[Logstash]] ← uses_buffer ← [[Redis]]
|
||||
- [[ELK Stack]] ← depends_on ← [[Logstash]]
|
||||
31
wiki/concepts/Management-Packs.md
Normal file
31
wiki/concepts/Management-Packs.md
Normal file
@@ -0,0 +1,31 @@
|
||||
---
|
||||
title: "Management Packs"
|
||||
type: concept
|
||||
tags: [Monitoring, OBM]
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
Management Packs 是 Micro Focus Operations Bridge Manager (OBM) 的管理包,用于定义监控策略、指标和阈值。
|
||||
|
||||
## Function
|
||||
- 定义监控间隔
|
||||
- 指定需要收集的指标
|
||||
- 配置数据收集范围(账户、服务、命名空间)
|
||||
- 设置阈值触发事件
|
||||
- 新实例自动部署监控策略
|
||||
|
||||
## Usage in OBM
|
||||
```yaml
|
||||
Management Pack Config:
|
||||
- Role ARN: arn:aws:iam::123456789012:role/OBM-Role
|
||||
- Account ID: target account
|
||||
- Namespaces: AWS/EC2, AWS/Lambda, etc.
|
||||
- Metrics: CPUUtilization, MemoryUtilization
|
||||
- Thresholds: 80% (warning), 90% (critical)
|
||||
- Frequency: 5 minutes
|
||||
```
|
||||
|
||||
## References
|
||||
- [[Operations Bridge Manager (OBM)]]
|
||||
- [[CTP Topic 8 Implementation of Cloud monitoring using Micro Focus Operations Bridge]]
|
||||
25
wiki/concepts/Node-Classes.md
Normal file
25
wiki/concepts/Node-Classes.md
Normal file
@@ -0,0 +1,25 @@
|
||||
---
|
||||
title: "Node Classes"
|
||||
type: concept
|
||||
tags: [Kubernetes, Karpenter, EC2, Instance-Configuration]
|
||||
sources: []
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
Node Classes 是 Karpenter 的核心组件,定义 EC2 实例的 provisioning 配置细节。
|
||||
|
||||
## Key Attributes
|
||||
- **子网**:实例部署的 subnet
|
||||
- **节点角色**:IAM 角色和策略
|
||||
- **AMI**:Amazon Machine Image(可使用 EKS 优化 AMI 或自定义 AMI)
|
||||
- **安全组**:实例附加的安全组
|
||||
|
||||
## Related Concepts
|
||||
- [[Node-Pools]]:定义调度约束和容量限制
|
||||
- [[Karpenter]]:使用 Node Classes 的 compute management tool
|
||||
- [[AMI]]:Amazon Machine Image
|
||||
|
||||
## Aliases
|
||||
- Node Classes
|
||||
- Karpenter Node Classes
|
||||
30
wiki/concepts/Node-Pool.md
Normal file
30
wiki/concepts/Node-Pool.md
Normal file
@@ -0,0 +1,30 @@
|
||||
---
|
||||
title: "Node Pool"
|
||||
description: EKS Auto Mode 的节点池概念,用于管理一组相同配置的 Kubernetes 节点
|
||||
type: concept
|
||||
tags:
|
||||
- AWS
|
||||
- EKS
|
||||
- Kubernetes
|
||||
---
|
||||
|
||||
## Definition
|
||||
Node Pool(节点池)是 EKS Auto Mode 中管理一组相同配置节点的概念。默认包含两个节点池:general purpose 和 system,还有一个节点类。
|
||||
|
||||
## Default Configuration
|
||||
- **General Purpose 节点池**:默认锁定 AMD64 架构,支持自定义节点池用于 Graviton 实例
|
||||
- **System 节点池**:带有 taint,系统插件需要相应的 tolerations
|
||||
- **节点类**:默认创建一个节点类
|
||||
|
||||
## Behavior
|
||||
- 默认节点池不可变,配置为零权重,允许自定义节点池被优先使用
|
||||
- 实例动态确定和自动整合
|
||||
- 优化计算成本
|
||||
|
||||
## Relationship
|
||||
- 由 [[Carpenter]] 管理
|
||||
- 运行 [[BottleRocket]] 操作系统
|
||||
- 属于 [[EKS-Auto-Mode]]
|
||||
|
||||
## References
|
||||
- [[public-cloud-learning-sessions-eks-optimization-part-3-of-3-introduction-to-eks-auto-mode]]
|
||||
30
wiki/concepts/Node-Pools.md
Normal file
30
wiki/concepts/Node-Pools.md
Normal file
@@ -0,0 +1,30 @@
|
||||
---
|
||||
title: "Node Pools"
|
||||
type: concept
|
||||
tags: [Kubernetes, Karpenter, Compute-Management]
|
||||
sources: []
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
Node Pools 是 Karpenter 的核心组件,定义 Kubernetes 集群中节点的调度约束和容量限制。
|
||||
|
||||
## Key Attributes
|
||||
- **调度约束**:定义哪些 Pod 可以调度到该节点池
|
||||
- **容量限制**:定义节点池的最大/最小节点数
|
||||
- **实例类型**:指定可使用的 EC2 实例类型
|
||||
- **可用区**:定义节点部署的可用区
|
||||
|
||||
## Use Cases
|
||||
- 单节点池:统一的工作负载
|
||||
- 混合节点池:计算+加速节点
|
||||
- 隔离节点池:基于成本、安全或多租户的隔离
|
||||
|
||||
## Related Concepts
|
||||
- [[Node-Classes]]:定义实例配置细节
|
||||
- [[Karpenter]]:使用 Node Pools 的 compute management tool
|
||||
- [[Consolidation-Policies]]:节点整合策略
|
||||
|
||||
## Aliases
|
||||
- Node Pools
|
||||
- Karpenter Node Pools
|
||||
27
wiki/concepts/OpenSearch.md
Normal file
27
wiki/concepts/OpenSearch.md
Normal file
@@ -0,0 +1,27 @@
|
||||
---
|
||||
title: "OpenSearch"
|
||||
type: concept
|
||||
tags: [Log-Analytics, AWS, Open-Source, Elasticsearch]
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Definition
|
||||
OpenSearch 是 AWS 的开源日志分析和搜索引擎,基于 Elasticsearch 和 Kibana 的开源分支,提供托管服务。
|
||||
|
||||
## Description
|
||||
OpenSearch 是 ELK Stack 的 AWS 托管替代方案,由 AWS 维护。与 Logz.io 等商业解决方案相比,OpenSearch 提供更低成本(约 $1,500/月 vs $4,000/月)和更高的可用性 SLA(99.9% vs 99.8%)。
|
||||
|
||||
## Features
|
||||
- 托管日志分析服务
|
||||
- 与 ELK Stack 兼容
|
||||
- 自动快照备份
|
||||
- 支持细粒度访问控制
|
||||
|
||||
## Cost Comparison(单农场、14天保留、每日 100GB)
|
||||
- Logz.io:约 $4,000/月
|
||||
- AWS OpenSearch:约 $1,500/月
|
||||
- 自托管 ELK:成本最低但维护量大
|
||||
|
||||
## Connections
|
||||
- [[OpenSearch]] ← extends ← [[ELK Stack]]
|
||||
- [[Log Analytics]] ← implements ← [[OpenSearch]]
|
||||
46
wiki/concepts/OpenTelemetry-Collector.md
Normal file
46
wiki/concepts/OpenTelemetry-Collector.md
Normal file
@@ -0,0 +1,46 @@
|
||||
---
|
||||
title: "OpenTelemetry Collector"
|
||||
type: concept
|
||||
tags:
|
||||
- OpenTelemetry
|
||||
- Collector
|
||||
- Data-Processing
|
||||
date: 2024-04-02
|
||||
---
|
||||
|
||||
## Definition
|
||||
OpenTelemetry Collector 是用于接收、处理和导出遥测数据的独立组件,作为数据管道在应用和后端存储之间进行数据标准化和转发。
|
||||
|
||||
## Architecture
|
||||
Collector 包含四大组件:
|
||||
|
||||
### Receivers(接收器)
|
||||
- AWS 特有接收器(ECS、EC2、EKS)
|
||||
- 开源接收器(Prometheus、OTLP、Fluent Bit)
|
||||
- 支持拉取(pull)和推送(push)模式
|
||||
|
||||
### Processors(处理器)
|
||||
- 数据过滤和转换
|
||||
- 批量处理和重试
|
||||
- 内存限流和队列管理
|
||||
|
||||
### Exporters(导出器)
|
||||
- AWS 原生:CloudWatch、X-Ray、OTLP
|
||||
- 开源:Prometheus、Jaeger、Zipkin
|
||||
- 第三方:Datadog、New Relic、Grafana
|
||||
|
||||
### Extensions(扩展)
|
||||
- SIGV 授权
|
||||
- 健康检查
|
||||
- zPages 调试
|
||||
- 内存配置
|
||||
|
||||
## Deployment Modes
|
||||
- **Agent 模式**:与应用部署在同一主机,收集本地数据
|
||||
- **Gateway 模式**:独立部署,收集多个 Agent 的数据
|
||||
|
||||
## Related Concepts
|
||||
- [[OpenTelemetry]]:父框架
|
||||
- [[ADOT]]:AWS 发行版,包含预配置 Collector
|
||||
- [[Fluent-Bit]]:日志收集器,可将数据转发至 Collector
|
||||
- [[OTLP]]:数据交换协议
|
||||
35
wiki/concepts/OpenTelemetry.md
Normal file
35
wiki/concepts/OpenTelemetry.md
Normal file
@@ -0,0 +1,35 @@
|
||||
---
|
||||
title: "OpenTelemetry"
|
||||
type: concept
|
||||
tags:
|
||||
- OpenTelemetry
|
||||
- Observability
|
||||
- Telemetry
|
||||
- Cloud-Native
|
||||
date: 2024-04-02
|
||||
---
|
||||
|
||||
## Definition
|
||||
OpenTelemetry(OTel)是厂商中立的遥测数据采集框架,提供统一的数据格式和跨语言 SDK,用于收集 Metrics、Logs、Traces 三种可观测性信号。
|
||||
|
||||
## Core Components
|
||||
- **SDKs**:支持 11 种编程语言的自动和手动 instrumentation 库
|
||||
- **Collector**:数据标准化和转发组件
|
||||
- **Protocol (OTLP)**:统一的遥测数据格式
|
||||
|
||||
## Key Capabilities
|
||||
- 厂商中立:不绑定特定后端监控系统
|
||||
- 自动 instrumentation:自动捕获常见框架和库的遥测数据
|
||||
- 统一数据格式:OTLP 标准化跨语言数据交换
|
||||
- 可扩展架构:支持自定义处理器和导出器
|
||||
|
||||
## Related Concepts
|
||||
- [[OpenTelemetry-Collector]]:数据收集和处理核心组件
|
||||
- [[ADOT]]:AWS 发行版
|
||||
- [[OTLP]]:OpenTelemetry Protocol 数据格式
|
||||
- [[可观测性三大支柱]]:Metrics、Logs、Traces
|
||||
|
||||
## Use Cases
|
||||
- 微服务架构的可观测性集成
|
||||
- 跨云厂商的监控数据统一采集
|
||||
- 统一日志、指标、追踪数据格式
|
||||
22
wiki/concepts/Pod-中断预算.md
Normal file
22
wiki/concepts/Pod-中断预算.md
Normal file
@@ -0,0 +1,22 @@
|
||||
---
|
||||
title: "Pod 中断预算"
|
||||
type: concept
|
||||
tags: [Kubernetes, EKS, High-Availability]
|
||||
---
|
||||
|
||||
## Description
|
||||
Pod 中断预算(Pod Disruption Budget,PDB)是 Kubernetes 机制,确保在自愿中断(如节点维护、升级)期间仍维持最少的 Pod 副本数,保证服务的可用性。
|
||||
|
||||
## Use Case
|
||||
在 EKS 中进行集群升级或节点维护时,PDB 确保应用始终保持最低服务水平。
|
||||
|
||||
## Related Concepts
|
||||
- [[EKS 可靠性]]
|
||||
- [[Pod 反亲和性]]
|
||||
|
||||
## Related Entities
|
||||
- [[EKS]]
|
||||
- [[Kubernetes]]
|
||||
|
||||
## Connections
|
||||
- [[Pod 中断预算]] ← 实现 [[EKS 可靠性]]
|
||||
23
wiki/concepts/Pod-反亲和性.md
Normal file
23
wiki/concepts/Pod-反亲和性.md
Normal file
@@ -0,0 +1,23 @@
|
||||
---
|
||||
title: "Pod 反亲和性"
|
||||
type: concept
|
||||
tags: [Kubernetes, EKS, High-Availability]
|
||||
---
|
||||
|
||||
## Description
|
||||
Pod 反亲和性(Pod Anti-Affinity)是一种 Kubernetes 调度策略,用于确保 Pod 不被调度到同一节点或同一可用区,从而提高应用的可用性和可靠性。
|
||||
|
||||
## Use Case
|
||||
在 EKS 中部署关键应用时,使用 Pod 反亲和性确保应用 Pod 分布在不同的节点和可用区,避免单点故障。
|
||||
|
||||
## Related Concepts
|
||||
- [[拓扑分布约束]]:Pod 反亲和性的更细粒度控制方式
|
||||
- [[HPA]]
|
||||
- [[VPA]]
|
||||
|
||||
## Related Entities
|
||||
- [[EKS]]
|
||||
- [[Kubernetes]]
|
||||
|
||||
## Connections
|
||||
- [[Pod 反亲和性]] ← 实现 [[EKS 可靠性]]
|
||||
20
wiki/concepts/SAML.md
Normal file
20
wiki/concepts/SAML.md
Normal file
@@ -0,0 +1,20 @@
|
||||
---
|
||||
title: "SAML"
|
||||
type: concept
|
||||
tags: []
|
||||
---
|
||||
|
||||
## Definition
|
||||
Security Assertion Markup Language,安全断言标记语言。基于 XML 的标准,用于在身份提供商和服务提供商之间交换身份验证和授权数据。
|
||||
|
||||
## Use Cases
|
||||
- 单点登录(SSO)
|
||||
- 多因素认证(MFA)集成
|
||||
- 跨域身份验证
|
||||
|
||||
## Related Entities
|
||||
- [[AppStream-2.0]]
|
||||
- [[AWS-Workspaces]]
|
||||
|
||||
## Related Concepts
|
||||
- [[BYOD]]
|
||||
37
wiki/concepts/Secrets-Management.md
Normal file
37
wiki/concepts/Secrets-Management.md
Normal file
@@ -0,0 +1,37 @@
|
||||
---
|
||||
title: "Secrets Management"
|
||||
type: concept
|
||||
tags: [security, devops, best-practices]
|
||||
sources: [ctp-topic-37-secrets-certificates-management, ctp-topic-62-aws-secrets-manager]
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Summary
|
||||
密钥管理是企业管理数字认证凭证(密码、API Token、加密密钥、证书)的系统性方法,确保应用服务、特权账户和 IT 生态系统中敏感信息的安全存储、访问控制和自动轮换。
|
||||
|
||||
## Definition
|
||||
管理数字认证凭证、密钥、密码、API 和 Token 等敏感信息的工具和方法,涵盖存储、访问控制、轮换、审计全生命周期。
|
||||
|
||||
## Core Components
|
||||
- **密钥存储**:集中化安全存储敏感信息
|
||||
- **访问控制**:基于身份的细粒度权限管理
|
||||
- **自动轮换**:定时自动更新密钥降低泄露风险
|
||||
- **审计日志**:记录所有访问和操作行为
|
||||
|
||||
## Implementation Patterns
|
||||
- **托管服务**:AWS Secrets Manager、Azure Key Vault、GCP Secret Manager
|
||||
- **自托管方案**:HashiCorp Vault(支持动态密钥、证书签名)
|
||||
- **特权访问管理**:CyberArk PAM、Micro Focus PAM
|
||||
|
||||
## Best Practices
|
||||
- 避免明文存储密钥
|
||||
- 实施最小权限原则
|
||||
- 启用自动轮换
|
||||
- 集中化密钥管理
|
||||
- 集成 CI/CD 流程
|
||||
|
||||
## Connections
|
||||
- [[Secrets Management]] ← 应用于 ← [[CI/CD]]
|
||||
- [[AWS Secrets Manager]] ← 实现 ← [[Secrets Management]]
|
||||
- [[HashiCorp Vault]] ← 实现 ← [[Secrets Management]]
|
||||
- [[Zero-Trust-Architecture]] ← 要求 ← [[Secrets Management]]
|
||||
29
wiki/concepts/Spot-Interruption.md
Normal file
29
wiki/concepts/Spot-Interruption.md
Normal file
@@ -0,0 +1,29 @@
|
||||
---
|
||||
title: "Spot Interruption"
|
||||
type: concept
|
||||
tags: [AWS, EC2, Cost-Optimization, High-Availability]
|
||||
sources: []
|
||||
last_updated: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
Spot Interruption 是 AWS EC2 Spot 实例的中断处理机制,Karpenter 原生集成 EventBridge 和 SQS 实现自动处理。
|
||||
|
||||
## Mechanism
|
||||
- **通知来源**:EventBridge 发送 Spot 中断、实例重平衡、健康事件、实例状态变更通知
|
||||
- **处理方式**:Karpenter 监听 SQS 队列,自动将工作负载迁移到其他实例
|
||||
- **无需额外组件**:Karpenter 原生支持,与 Node Termination Handler 不同
|
||||
|
||||
## Benefits
|
||||
- 降低 Spot 实例使用成本(可达 90% 折扣)
|
||||
- 自动处理中断,提高工作负载可用性
|
||||
- 简化架构,无需额外管理组件
|
||||
|
||||
## Related Concepts
|
||||
- [[Karpenter]]:原生支持 Spot 中断处理
|
||||
- [[EventBridge]]:AWS 事件总线服务
|
||||
- [[SQS]]:AWS 简单队列服务
|
||||
|
||||
## Aliases
|
||||
- Spot Interruption Handling
|
||||
- Spot Instance Interruption
|
||||
21
wiki/concepts/VDI.md
Normal file
21
wiki/concepts/VDI.md
Normal file
@@ -0,0 +1,21 @@
|
||||
---
|
||||
title: "VDI"
|
||||
type: concept
|
||||
tags: []
|
||||
---
|
||||
|
||||
## Definition
|
||||
Virtual Desktop Infrastructure,虚拟桌面基础设施。通过远程桌面协议(如 RDP、WSP)提供虚拟计算环境,用户可以在任何设备上访问托管的虚拟桌面。
|
||||
|
||||
## Use Cases
|
||||
- 远程办公和混合工作模式
|
||||
- BYOD(自带设备)安全访问企业资源
|
||||
- 标准化桌面管理和安全控制
|
||||
|
||||
## Related Concepts
|
||||
- [[持久化桌面]]:数据和应用设置在会话间保留
|
||||
- [[非持久化桌面]]:每次登录分配新桌面
|
||||
|
||||
## Related Entities
|
||||
- [[AWS-Workspaces]]
|
||||
- [[AppStream-2.0]]
|
||||
22
wiki/concepts/VPA.md
Normal file
22
wiki/concepts/VPA.md
Normal file
@@ -0,0 +1,22 @@
|
||||
---
|
||||
title: "VPA"
|
||||
type: concept
|
||||
tags: [Kubernetes, EKS, Auto-Scaling]
|
||||
---
|
||||
|
||||
## Description
|
||||
VPA(Vertical Pod Autoscaler)是 Kubernetes 的垂直 Pod 自动扩缩容功能,根据实际使用情况自动调整 Pod 的资源请求(CPU、内存),但调整时会重启 Pod。
|
||||
|
||||
## Note
|
||||
与 HPA 不同,VPA 调整的是单个 Pod 的资源大小而非副本数。
|
||||
|
||||
## Related Concepts
|
||||
- [[HPA]]
|
||||
- [[EKS 可靠性]]
|
||||
|
||||
## Related Entities
|
||||
- [[EKS]]
|
||||
- [[Kubernetes]]
|
||||
|
||||
## Connections
|
||||
- [[VPA]] ← 实现 [[EKS 可靠性]]
|
||||
16
wiki/concepts/WSP-Protocol.md
Normal file
16
wiki/concepts/WSP-Protocol.md
Normal file
@@ -0,0 +1,16 @@
|
||||
---
|
||||
title: "WSP Protocol"
|
||||
type: concept
|
||||
tags: []
|
||||
---
|
||||
|
||||
## Definition
|
||||
Workspaces Streaming Protocol,AWS Workspaces 专用的桌面流协议。专为高延迟网络环境设计,优化远程桌面的用户体验。
|
||||
|
||||
## Characteristics
|
||||
- 专为高延迟网络优化
|
||||
- 支持剪贴板、摄像头、智能卡等外设
|
||||
- 适用于远程工作和分布式团队
|
||||
|
||||
## Related Entities
|
||||
- [[AWS-Workspaces]]
|
||||
25
wiki/concepts/dm-verity.md
Normal file
25
wiki/concepts/dm-verity.md
Normal file
@@ -0,0 +1,25 @@
|
||||
---
|
||||
title: "dm-verity"
|
||||
type: concept
|
||||
tags: [Linux-Kernel, Security, Filesystem]
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
dm-verity(device-mapper verity)是 Linux 内核子系统,用于验证块设备的完整性,通过 cryptographic hash 树实现只读文件系统的完整性保护。
|
||||
|
||||
## How It Works
|
||||
- 在块设备上构建 hash 树结构
|
||||
- 每个块的数据 hash 与上一级 hash 比对
|
||||
- 根 hash 存储在信任的存储位置
|
||||
- 任何块内容变化都会导致验证失败
|
||||
|
||||
## Use Cases
|
||||
- 防止根文件系统被篡改
|
||||
- 确保容器镜像完整性
|
||||
- 安全启动链的一部分
|
||||
|
||||
## Related Concepts
|
||||
- [[Bottlerocket-OS]] — 使用 dm-verity 验证根文件系统
|
||||
- [[Secure-Boot]] — 安全启动机制
|
||||
- [[Root-Filesystem]] — 根文件系统保护
|
||||
51
wiki/concepts/可观测性三大支柱.md
Normal file
51
wiki/concepts/可观测性三大支柱.md
Normal file
@@ -0,0 +1,51 @@
|
||||
---
|
||||
title: "可观测性三大支柱"
|
||||
type: concept
|
||||
tags:
|
||||
- Observability
|
||||
- Metrics
|
||||
- Logs
|
||||
- Traces
|
||||
date: 2024-04-02
|
||||
---
|
||||
|
||||
## Definition
|
||||
可观测性三大支柱是系统可观测性的三个核心信号:Metrics(指标)、Logs(日志)、Traces(追踪),它们相互关联共同提供系统内部状态的可见性。
|
||||
|
||||
## Three Pillars
|
||||
|
||||
### Metrics(指标)
|
||||
- 聚合的源统计数据
|
||||
- 长时间序列的数值测量
|
||||
- 用于监控、告警和趋势分析
|
||||
- 示例:CPU 使用率、请求延迟、错误率
|
||||
|
||||
### Logs(日志)
|
||||
- 事件的时间戳记录
|
||||
- 详细的事件描述信息
|
||||
- 用于问题根因分析和调试
|
||||
- 示例:应用错误日志、访问日志
|
||||
|
||||
### Traces(追踪)
|
||||
- 请求在分布式系统中的完整路径
|
||||
- 包含多个 Span(跨度)形成的调用链
|
||||
- 用于理解系统行为和性能瓶颈
|
||||
- 示例:用户请求从 API Gateway → Service A → Service B → Database
|
||||
|
||||
## Trace Span
|
||||
追踪中的一个单元,包含:
|
||||
- 开始时间(Start Time)
|
||||
- 持续时间(Duration)
|
||||
- 元数据(Metadata)
|
||||
- 关联的日志
|
||||
|
||||
## Integration
|
||||
- OpenTelemetry 统一收集这三种信号
|
||||
- 通过关联分析实现端到端可观测性
|
||||
- Grafana/OpenSearch 可视化展示
|
||||
|
||||
## Related Concepts
|
||||
- [[OpenTelemetry]]:统一采集框架
|
||||
- [[OpenTelemetry-Collector]]:数据收集组件
|
||||
- [[ADOT]]:AWS 发行版
|
||||
- [[Fluent-Bit]]:日志收集器
|
||||
22
wiki/concepts/拓扑分布约束.md
Normal file
22
wiki/concepts/拓扑分布约束.md
Normal file
@@ -0,0 +1,22 @@
|
||||
---
|
||||
title: "拓扑分布约束"
|
||||
type: concept
|
||||
tags: [Kubernetes, EKS, High-Availability]
|
||||
---
|
||||
|
||||
## Description
|
||||
拓扑分布约束(Topology Spread Constraints)是 Kubernetes 提供的 Pod 分布控制机制,比 Pod 反亲和性更细粒度,可以控制 Pod 在可用区、节点等拓扑域的分布。
|
||||
|
||||
## Use Case
|
||||
在 EKS 中确保工作负载在多个可用区间均匀分布,实现高可用性。
|
||||
|
||||
## Related Concepts
|
||||
- [[Pod 反亲和性]]:更简单的 Pod 分布策略
|
||||
- [[EKS 可靠性]]
|
||||
|
||||
## Related Entities
|
||||
- [[EKS]]
|
||||
- [[Kubernetes]]
|
||||
|
||||
## Connections
|
||||
- [[拓扑分布约束]] ← 实现 [[EKS 可靠性]]
|
||||
26
wiki/concepts/持久化桌面.md
Normal file
26
wiki/concepts/持久化桌面.md
Normal file
@@ -0,0 +1,26 @@
|
||||
---
|
||||
title: "持久化桌面"
|
||||
type: concept
|
||||
tags: []
|
||||
---
|
||||
|
||||
## Definition
|
||||
每个用户独享虚拟机实例,应用状态和设置在会话之间保持。下次登录时,用户看到相同的工作环境和个人文件。
|
||||
|
||||
## Characteristics
|
||||
- 一对一实例管理
|
||||
- 用户数据和应用状态持久化
|
||||
- 适合需要个性化环境的知识工作者
|
||||
- AWS Workspaces 是典型实现
|
||||
|
||||
## Use Cases
|
||||
- 长期办公的远程员工
|
||||
- 需要保留工作环境的开发人员
|
||||
- 需要访问企业应用的 BYOD 用户
|
||||
|
||||
## Related Entities
|
||||
- [[AWS-Workspaces]]
|
||||
|
||||
## Related Concepts
|
||||
- [[VDI]]
|
||||
- [[非持久化桌面]]
|
||||
21
wiki/concepts/探针.md
Normal file
21
wiki/concepts/探针.md
Normal file
@@ -0,0 +1,21 @@
|
||||
---
|
||||
title: "探针"
|
||||
type: concept
|
||||
tags: [Kubernetes, EKS, Health-Check]
|
||||
---
|
||||
|
||||
## Description
|
||||
探针(Probe)是 Kubernetes 用于检测 Pod 健康状态的功能,包含三种类型:
|
||||
- **Liveness Probe**:检测容器是否存活,失败时重启容器
|
||||
- **Readiness Probe**:检测容器是否就绪,失败时从 Service 移除流量
|
||||
- **Startup Probe**:检测容器是否启动完成,失败时重启容器(适用于启动慢的应用)
|
||||
|
||||
## Related Concepts
|
||||
- [[EKS 可靠性]]
|
||||
|
||||
## Related Entities
|
||||
- [[EKS]]
|
||||
- [[Kubernetes]]
|
||||
|
||||
## Connections
|
||||
- [[探针]] ← 实现 [[EKS 可靠性]]
|
||||
32
wiki/concepts/最小权限原则.md
Normal file
32
wiki/concepts/最小权限原则.md
Normal file
@@ -0,0 +1,32 @@
|
||||
---
|
||||
title: "最小权限原则"
|
||||
type: concept
|
||||
tags: [Security, IAM, Best-Practice]
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
最小权限原则(Least Privilege)是安全最佳实践,只授予完成任务所需的最小权限。
|
||||
|
||||
## Core Concept
|
||||
> "We only want to allow the access that is strictly required."
|
||||
|
||||
只授予完成任务所需的最小权限,降低权限滥用和数据泄露风险。
|
||||
|
||||
## Implementation
|
||||
- 从空白策略开始,逐步添加所需权限
|
||||
- 定期审查和调整权限
|
||||
- 使用资源级别限制特定资源而非广泛权限
|
||||
- 避免使用通配符(*)
|
||||
|
||||
## Related Concepts
|
||||
- [[IAM-策略]]: 最小权限的应用对象
|
||||
- [[IAM-用户]]: 需要最小权限管理的实体
|
||||
|
||||
## Role in Cloud Security
|
||||
- 降低数据泄露影响范围
|
||||
- 限制内部威胁
|
||||
- 满足合规要求(PCI-DSS、HIPAA、GDPR)
|
||||
|
||||
## Connections
|
||||
- [[最小权限原则]] ← guides ← [[IAM-策略]]
|
||||
43
wiki/concepts/联合访问.md
Normal file
43
wiki/concepts/联合访问.md
Normal file
@@ -0,0 +1,43 @@
|
||||
---
|
||||
title: "联合访问"
|
||||
type: concept
|
||||
tags: [AWS, IAM, Federation, Security]
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Definition
|
||||
联合访问是通过外部身份提供商(如 Active Directory)映射 IAM 角色实现 AWS 访问的方式。
|
||||
|
||||
## Core Concept
|
||||
> "Federated users log in via their organization's AD, which maps to an IAM role."
|
||||
|
||||
联合用户通过组织的 AD 登录,AD 组映射到 IAM 角色,角色授予相应的 AWS 访问权限。
|
||||
|
||||
## Workflow
|
||||
1. 用户通过 AD 凭证登录
|
||||
2. AD 组映射到 IAM 角色
|
||||
3. 用户 assum 角色获取临时 AWS 凭证
|
||||
4. 通过 CLI 工具(如 PFSSO)访问 AWS
|
||||
|
||||
## Components
|
||||
- **Active Directory**:外部身份提供商
|
||||
- **AD 组**:映射到 IAM 角色的组
|
||||
- **IAM 角色**:接收 AD 映射的角色
|
||||
- **PFSSO**:命令行联合访问工具
|
||||
|
||||
## Why Federation
|
||||
- 集中管理用户身份(无需单独管理 IAM 用户)
|
||||
- 员工离职后自动失去访问权限
|
||||
- 单一登录入口
|
||||
|
||||
## Best Practice
|
||||
Federation 是用户管理的首选方法,IAM 用户仅用于服务账号。
|
||||
|
||||
## Related Concepts
|
||||
- [[IAM-角色]]: 联合访问的目标角色
|
||||
- [[PFSSO]]: 命令行联合访问工具
|
||||
- [[Active-Directory]]: 外部身份提供商
|
||||
|
||||
## Connections
|
||||
- [[AD]] ← maps_to_role ← [[IAM-角色]]
|
||||
- [[联合访问]] ← requires ← [[PFSSO]]
|
||||
27
wiki/concepts/非持久化桌面.md
Normal file
27
wiki/concepts/非持久化桌面.md
Normal file
@@ -0,0 +1,27 @@
|
||||
---
|
||||
title: "非持久化桌面"
|
||||
type: concept
|
||||
tags: []
|
||||
---
|
||||
|
||||
## Definition
|
||||
每次登录时分配新的桌面环境,会话结束后数据不被保留。多个用户可以共享同一实例,实现资源复用。
|
||||
|
||||
## Characteristics
|
||||
- 多对一实例管理(多租户)
|
||||
- 会话结束后数据清除
|
||||
- 通过应用和存储连接器可实现选择性持久化
|
||||
- 适合共享资源场景
|
||||
|
||||
## Use Cases
|
||||
- 实验室环境
|
||||
- 培训场景
|
||||
- 临时访客
|
||||
- 任务型工作者
|
||||
|
||||
## Related Entities
|
||||
- [[AppStream-2.0]]
|
||||
|
||||
## Related Concepts
|
||||
- [[VDI]]
|
||||
- [[持久化桌面]]
|
||||
Reference in New Issue
Block a user