ingmarioalberto · December 1, 2022 18:35
diff --git a/TLF_LFS162x.txt b/TLF_LFS162x.txt
 TLF (LFS162x) (Prof. Gourav Shah)
 Introduction to DevOps and Site Reliability Engineering
 DevOps = Developer & Operations (Areas)
 SRE = Site Reliability Engineer
 ---
 Needs: Bridge Gap between ( Faster Dev vs Waterfall practices )
 DevOps via:
 	( Modern principles
 	Practices
 	Automation tools ) --PATH--> Ops into Agile Area => Faster software delivery.
 ---
 cloud ~transforming~> organizations by providing:
 	-elastic
 	-scalable
 	-HA
 	-Multiregional infrastructure
 ---
 Learn:
 	* Containers are revolut. software delivery
 	* Intrastructure as code & tools
 	* Continuous Integration and Continuos Delivery differences.
 	* Observability Systems
 ----------
 Pre DevOps Era
 	Puppet as Config management
 	Centralized management system
 	Write infrastructure as code and publish to centralized server.
 	Google: (~2008)
 		Resource isolation & namespaces = cgroups
 ---
 (~2009)
 Andrew Clay & Patrick Debois (Belgium) =>Agile Systems Administrator Group
 =>DevOps term (2009).
 Puppet -> Chef
 ---
 (~2010-2015) = Devops Chaos Era
 	(~2011): Hudson -forked-> Jenkins
 	(~2012):
 		Ansible
 		DotCloud opensourced Docker
 	(~2014-2015:
 		orchestration engines:
 			* Google opensourced Kubernetes, run infrastructure over Docker in production envs.
 			* Docker released Swarm
 			* Apache Mesos released Marathon
 	(~2015):
 		Configuration management:
 			Puppet
 			Chef
 			Saltstack
 			Ansible
 		Continuous Integration:
 			* Jenkins
 			* Others like: Bamboo, Go, Travis CI, Circle CI
 ---
 (~2016-2018) = Devops Maturity years
 	(~2016): Kubernetes turned into the most reliable 
 	(~2018):
 		Mesos and Docker became Kubernetes friendly
 		Ansible became as prefered config mgmt
 ---
 (~2019-2020)
 	{reli,avail,scal}ability & security
 	five pillars of devops: 
 		* the cloud
 			lead by: AWS, Azure, GCP
 		* containers: 
 			Kubernetes and Docker are standard now
 			Kubernetes:
 				ability to scale, load balance, HA, service discovery
 		* infrastructure as code (IaaC):
 			Ansible can do it
 			Terraform is specialized.
 			Others: 
 				Docker Compose
 				Kubernetes scripts
 		* CI / CD:
 			lead by: 
 				Jenkins with Spinnaker
 		* Observability
 				Based on Collecting:
 					Logs, Metrics,	Traces.
 				Guess the state of the system
 			Logs:
 				Elasticsearch, Logstash, Kibana
 			Metrics:
 				Prometheus: Time Series database
 				Grafana: Visualizer
 			Traces:
 				Jaeger
 				Zipkin
 ---
 	(Present?)
 	Kubernetes is the platform on top oh which things are built.
 	Custom resource definitions (Service Mesh):		
 		Istio & Linkerd:
 			Telemetry example: Routing traffic
 	Jenkins X
 		CI/CD for cloud native apps over kubernetes
 		integrates with Jenkins, Github with gitops
 	Tekton forked from Knative lead by Google and IBM allows to CI/CD within kubernetes using YAML, features like blue green, canary deployments, gitops.		
 ---
 	Deepen/Review later:
 		Pipeline as a Code
 		Spinnaker: Hybrid cloud deploment tool
 		Custom resource definitions
 		Service Mesh:
 			Istio 
 			Linkerd
 		Gitops	
 		Revisar curso:
 			The Linux Foundation and CDF by the name DevOps and SRE Fundamentals: Implementing Continuous Delivery, with the course code LFS261 (Jenkins and Spinaker)
 ----------3
 Cloud - Why Should You Bother?
 	Site Reliability/DevOps Engineering: SR/DE
 	9 reasons to bother about cloud:
 		* utility computing model:
 			buy the service instead of building
 		* cloud = access to global infrastructure
 		* scalable platform
 		* OPEX: Operational Expenses
 				vs 
 			CAPEX: Capital Expenses 
 		* Dynamic Capacity Provisioning and Scalability
 			Scalability ~=
 				Resilence
 				Fault Tolerance
 			High Availability ~=
 				Multiple geolocations
 		* Business continuity
 				Multiple geolocations
 		* Managed services
 				Backups, Snapshots, HA, Scalability everything is done by the cloud platform.
 		* Automation out of the box
 			Terraform -------
 			Ansible----------+\ IaaC
 			Cloudformation---+/
 			Heat Templates---
 ---
 A Brief History of Cloud:
 	John McCarthy (1961), time-sharing model will result in computing power sold as utility.
 	(60s-70s): 
 		Distributed computers internconnected.
 	(~1991):
 		Tim Berners -> WWW
 	(~1999):
 		VMWare -> Virtualization
 	(~2006):
 		AWS -> IaaS
 	(~2008):
 		AWS -> VPCloud, Storage Elastic Block, Load Balancers and Auto-scaling
 		Google -> Google compuite engine
 	(~2009):
 		Microsoft -> Azure
 ---
 Types of Cloud:
 	IaaS : Infrastructure as a Service
 		Complete control over servers, storage, network, databases.
 		Receiving the Infrastructure as ready working pieces
 		ex: AWS,GCP,Azute
 	PaaS : Platform as a Service
 		Microsoft Azure Services, 
 		AWS Services,
 		AWS Elastic Beanstalk 
 		Red Hat Openshift will automatically:
 			Takes you repo
 			arrange it into a docker image
 			deploy at scale
 			deploy databases
 			load balancers
 			network config
 	SaaS : Software as a Service
 		Gmail, Dropbox, 
 ---
 Cloud Deployment Models:
 	Public: 
 		Shared Infrastructure available for anyone globally
 		ex: AWS, GCP, DigitalOcean, …
 	Private
 		Something we build inside, in our own datacenter, with the ability for self-service or self-provisioning.
 		Openstack help to build our own private cloud platform
 	Hybrid
 		When you want some extra power (burst) or need to host something on another coutry, but keep other services on my own private cloud,
 ---
 Major Cloud Services
 	Compute
 	Databases
 		SQL, NoSQL
 	Storage
 		Block storage
 		Object based storage
 	Network services
 		virtual private cloud ~= network config
 		Network ACLs ~= Firewall at network level
 		Subnets
 		Private Subnets
 		Custom IPs
 	User control
 		IAM: Identity and Access Management
 	DevOps services
 		Service management
 		Push notifications services
 		Analytics services 
 		Routing, DNS, Email, etc			

 	Deepen/Review later:
 			Block storage
 			Object based storage
 ---
 Demos: Building Scalable, High Available Infrastructure with AWS		
 Virtual Private Cloud: VPC
 Availability Zones (AZ) ~= Datacenters
 Auto Scaling Groups: ASG


 VPC:
 	Define 
 		Network ACLs: (Firewall rules)
 			what go In and/or Out
 		security groups: (firewall at software level)
 		EC2 instances ~= Servers
 		RDS = Relational Database Service
 000000000111111111122222222223333333333444		
 123456789012345678901234567890123456789012

 ///////////////Network ACLs///////////////
 ╔════════════════════════════════════════╗ 
 ║           [ Load Balancer ]            ║
 ║                /      \                ║
 ╟------------======S.G.======------------╢
 ║  . . . . . . . . .║. . . . . . .ASG    ║
 ║ .┌---┐            ║  ┌---┐      .      ║
 ║ .│EC²│            ║  │EC²│      .      ║ 
 ║ .└-┬-┘            ║  └-┬-┘      .      ║
 ║  . . . . . . . . .║. . . . . . .       ║
 ║    |   DB(RDS)    ║    |   DB          ║
 ║    └---▊|         ║    └---▊|          ║
 ║         ^_____repl║ica_____^           ║
 ║                   ║                    ║
 ║                   ║                    ║
 ║       AZ1      <-Netw->    AZ2         ║
 ╚═══════════════════╩════════════════════╝
 									 VPC

 Generate This automatically with:
 	Cloudformation

 ----------------------------------------------------
 CI/CD - Why Should You Bother?
 It's all about reliability no matter the features
 	available
 	scalable
 	
 	bugs affects reliability ~= downtimes
 			rework
 	minimize deffects ~= Continuous and Automated Feedback
 ----------------------------------------------------
 What Is Continuous Integration?	
 	simple testing can prevent most critical of failures.
 When:
 	Whenever you integrate code to the mainline

 Continuous Integration = Integration + Build + Unit Tests + Static Checkers  + {Integration+Acceptance+Load+Compliance} tests

 Agile = Smaller Batches + Integrate Often
 so:
 	CI = Code merged into master is reliable
 ----------------------------------------------------
 CI Tool:
 	Ability to create jobs to run build, tests, packaging by integrating with some tools
 	Connect those jobs to create a sequence of execution pipelines.
 	Run pipelines automatically
 	Trigger pipelines based on events
 Tools
 	Jenkins, Go, Bamboo, 
 	Travis CI, Circle CI	
 ----------------------------------------------------
 Continuous Deployment and Continuous Delivery
 Build
 	->Test
 		->Package
 			->Deploy to DEV
 				->Deploy to Stage
 					->Ask for permission
 						->Deploy to Prod
 Why?
 	Controlled and safer
 	Stress free deployments
 	Allows for manual and exploratory testing
 	Could accommodate hybrid deployments
 	Scheduled releases
 ----------------------------------------------------
 CD = Build->Test->Package->Deploy to {DEV->STAGE->ReleaseManager->PROD}
 CD -> Controlled and Safer, Stress free deployments, , allows manual or exploratory testing, hybrid deployments (schema updates + auto deploy), Schedules releases, 

 CI/CD Principles and Practices
 Agile <--> CD
 Build=Automated + Quick
 Test should be automated:
 	Test Pyramid:
 	(Base)->Automated{Unit->Component->Service->API->UI->Manual}Test
 	The more at the right (top of the pyramid) more expensive 
 Deployments=Automated
 Practices:
  Revision control:
 	  Application Code
 		Build logic (pom.xml, package.json, makefiles,etc)
 		Configurations
 		CI Pipelines (jenkinsFiles)
 		Artifact build and packaging logic (dockerfiles)
 		IaaCode (ansible)
 		Deployment logic (Kubernetes deployment files)
 	Development workflow: How fast to deliver
 	Automated Deployment Pipelines (Spinnaker)
 Summary:
 	CD:
 		Setup revision control system (git+github)
 		Define development workflow + Code review
 		Setup CI Platform
 		Create CI Pipelines: Build, Test
 		Build and Manage artefacts: Packaging, Jobs
 		Setup deployment tools and create deployment pipelines
 		
 ╔═══════╦═════════╗ 
 ║       ║    ╨    ║
 ╠       ╬
 ╚═══════╩═════════╝
	TLF (LFS162x) (Prof. Gourav Shah)
	Introduction to DevOps and Site Reliability Engineering
	DevOps = Developer & Operations (Areas)
	SRE = Site Reliability Engineer
	---
	Needs: Bridge Gap between ( Faster Dev vs Waterfall practices )
	DevOps via:
	( Modern principles
	Practices
	Automation tools ) --PATH--> Ops into Agile Area => Faster software delivery.
	---
	cloud ~transforming~> organizations by providing:
	-elastic
	-scalable
	-HA
	-Multiregional infrastructure
	---
	Learn:
	* Containers are revolut. software delivery
	* Intrastructure as code & tools
	* Continuous Integration and Continuos Delivery differences.
	* Observability Systems
	----------
	Pre DevOps Era
	Puppet as Config management
	Centralized management system
	Write infrastructure as code and publish to centralized server.
	Google: (~2008)
	Resource isolation & namespaces = cgroups
	---
	(~2009)
	Andrew Clay & Patrick Debois (Belgium) =>Agile Systems Administrator Group
	=>DevOps term (2009).
	Puppet -> Chef
	---
	(~2010-2015) = Devops Chaos Era
	(~2011): Hudson -forked-> Jenkins
	(~2012):
	Ansible
	DotCloud opensourced Docker
	(~2014-2015:
	orchestration engines:
	* Google opensourced Kubernetes, run infrastructure over Docker in production envs.
	* Docker released Swarm
	* Apache Mesos released Marathon
	(~2015):
	Configuration management:
	Puppet
	Chef
	Saltstack
	Ansible
	Continuous Integration:
	* Jenkins
	* Others like: Bamboo, Go, Travis CI, Circle CI
	---
	(~2016-2018) = Devops Maturity years
	(~2016): Kubernetes turned into the most reliable
	(~2018):
	Mesos and Docker became Kubernetes friendly
	Ansible became as prefered config mgmt
	---
	(~2019-2020)
	{reli,avail,scal}ability & security
	five pillars of devops:
	* the cloud
	lead by: AWS, Azure, GCP
	* containers:
	Kubernetes and Docker are standard now
	Kubernetes:
	ability to scale, load balance, HA, service discovery
	* infrastructure as code (IaaC):
	Ansible can do it
	Terraform is specialized.
	Others:
	Docker Compose
	Kubernetes scripts
	* CI / CD:
	lead by:
	Jenkins with Spinnaker
	* Observability
	Based on Collecting:
	Logs, Metrics, Traces.
	Guess the state of the system
	Logs:
	Elasticsearch, Logstash, Kibana
	Metrics:
	Prometheus: Time Series database
	Grafana: Visualizer
	Traces:
	Jaeger
	Zipkin
	---
	(Present?)
	Kubernetes is the platform on top oh which things are built.
	Custom resource definitions (Service Mesh):
	Istio & Linkerd:
	Telemetry example: Routing traffic
	Jenkins X
	CI/CD for cloud native apps over kubernetes
	integrates with Jenkins, Github with gitops
	Tekton forked from Knative lead by Google and IBM allows to CI/CD within kubernetes using YAML, features like blue green, canary deployments, gitops.
	---
	Deepen/Review later:
	Pipeline as a Code
	Spinnaker: Hybrid cloud deploment tool
	Custom resource definitions
	Service Mesh:
	Istio
	Linkerd
	Gitops
	Revisar curso:
	The Linux Foundation and CDF by the name DevOps and SRE Fundamentals: Implementing Continuous Delivery, with the course code LFS261 (Jenkins and Spinaker)
	----------3
	Cloud - Why Should You Bother?
	Site Reliability/DevOps Engineering: SR/DE
	9 reasons to bother about cloud:
	* utility computing model:
	buy the service instead of building
	* cloud = access to global infrastructure
	* scalable platform
	* OPEX: Operational Expenses
	vs
	CAPEX: Capital Expenses
	* Dynamic Capacity Provisioning and Scalability
	Scalability ~=
	Resilence
	Fault Tolerance
	High Availability ~=
	Multiple geolocations
	* Business continuity
	Multiple geolocations
	* Managed services
	Backups, Snapshots, HA, Scalability everything is done by the cloud platform.
	* Automation out of the box
	Terraform -------
	Ansible----------+\ IaaC
	Cloudformation---+/
	Heat Templates---
	---
	A Brief History of Cloud:
	John McCarthy (1961), time-sharing model will result in computing power sold as utility.
	(60s-70s):
	Distributed computers internconnected.
	(~1991):
	Tim Berners -> WWW
	(~1999):
	VMWare -> Virtualization
	(~2006):
	AWS -> IaaS
	(~2008):
	AWS -> VPCloud, Storage Elastic Block, Load Balancers and Auto-scaling
	Google -> Google compuite engine
	(~2009):
	Microsoft -> Azure
	---
	Types of Cloud:
	IaaS : Infrastructure as a Service
	Complete control over servers, storage, network, databases.
	Receiving the Infrastructure as ready working pieces
	ex: AWS,GCP,Azute
	PaaS : Platform as a Service
	Microsoft Azure Services,
	AWS Services,
	AWS Elastic Beanstalk
	Red Hat Openshift will automatically:
	Takes you repo
	arrange it into a docker image
	deploy at scale
	deploy databases
	load balancers
	network config
	SaaS : Software as a Service
	Gmail, Dropbox,
	---
	Cloud Deployment Models:
	Public:
	Shared Infrastructure available for anyone globally
	ex: AWS, GCP, DigitalOcean, …
	Private
	Something we build inside, in our own datacenter, with the ability for self-service or self-provisioning.
	Openstack help to build our own private cloud platform
	Hybrid
	When you want some extra power (burst) or need to host something on another coutry, but keep other services on my own private cloud,
	---
	Major Cloud Services
	Compute
	Databases
	SQL, NoSQL
	Storage
	Block storage
	Object based storage
	Network services
	virtual private cloud ~= network config
	Network ACLs ~= Firewall at network level
	Subnets
	Private Subnets
	Custom IPs
	User control
	IAM: Identity and Access Management
	DevOps services
	Service management
	Push notifications services
	Analytics services
	Routing, DNS, Email, etc

	Deepen/Review later:
	Block storage
	Object based storage
	---
	Demos: Building Scalable, High Available Infrastructure with AWS
	Virtual Private Cloud: VPC
	Availability Zones (AZ) ~= Datacenters
	Auto Scaling Groups: ASG


	VPC:
	Define
	Network ACLs: (Firewall rules)
	what go In and/or Out
	security groups: (firewall at software level)
	EC2 instances ~= Servers
	RDS = Relational Database Service
	000000000111111111122222222223333333333444
	123456789012345678901234567890123456789012

	///////////////Network ACLs///////////////
	╔════════════════════════════════════════╗
	║ [ Load Balancer ] ║
	║ / \ ║
	╟------------======S.G.======------------╢
	║ . . . . . . . . .║. . . . . . .ASG ║
	║ .┌---┐ ║ ┌---┐ . ║
	║ .│EC²│ ║ │EC²│ . ║
	║ .└-┬-┘ ║ └-┬-┘ . ║
	║ . . . . . . . . .║. . . . . . . ║
	║ \| DB(RDS) ║ \| DB ║
	║ └---▊\| ║ └---▊\| ║
	║ ^_____repl║ica_____^ ║
	║ ║ ║
	║ ║ ║
	║ AZ1 <-Netw-> AZ2 ║
	╚═══════════════════╩════════════════════╝
	VPC

	Generate This automatically with:
	Cloudformation

	----------------------------------------------------
	CI/CD - Why Should You Bother?
	It's all about reliability no matter the features
	available
	scalable

	bugs affects reliability ~= downtimes
	rework
	minimize deffects ~= Continuous and Automated Feedback
	----------------------------------------------------
	What Is Continuous Integration?
	simple testing can prevent most critical of failures.
	When:
	Whenever you integrate code to the mainline

	Continuous Integration = Integration + Build + Unit Tests + Static Checkers + {Integration+Acceptance+Load+Compliance} tests

	Agile = Smaller Batches + Integrate Often
	so:
	CI = Code merged into master is reliable
	----------------------------------------------------
	CI Tool:
	Ability to create jobs to run build, tests, packaging by integrating with some tools
	Connect those jobs to create a sequence of execution pipelines.
	Run pipelines automatically
	Trigger pipelines based on events
	Tools
	Jenkins, Go, Bamboo,
	Travis CI, Circle CI
	----------------------------------------------------
	Continuous Deployment and Continuous Delivery
	Build
	->Test
	->Package
	->Deploy to DEV
	->Deploy to Stage
	->Ask for permission
	->Deploy to Prod
	Why?
	Controlled and safer
	Stress free deployments
	Allows for manual and exploratory testing
	Could accommodate hybrid deployments
	Scheduled releases
	----------------------------------------------------
	CD = Build->Test->Package->Deploy to {DEV->STAGE->ReleaseManager->PROD}
	CD -> Controlled and Safer, Stress free deployments, , allows manual or exploratory testing, hybrid deployments (schema updates + auto deploy), Schedules releases,

	CI/CD Principles and Practices
	Agile <--> CD
	Build=Automated + Quick
	Test should be automated:
	Test Pyramid:
	(Base)->Automated{Unit->Component->Service->API->UI->Manual}Test
	The more at the right (top of the pyramid) more expensive
	Deployments=Automated
	Practices:
	Revision control:
	Application Code
	Build logic (pom.xml, package.json, makefiles,etc)
	Configurations
	CI Pipelines (jenkinsFiles)
	Artifact build and packaging logic (dockerfiles)
	IaaCode (ansible)
	Deployment logic (Kubernetes deployment files)
	Development workflow: How fast to deliver
	Automated Deployment Pipelines (Spinnaker)
	Summary:
	CD:
	Setup revision control system (git+github)
	Define development workflow + Code review
	Setup CI Platform
	Create CI Pipelines: Build, Test
	Build and Manage artefacts: Packaging, Jobs
	Setup deployment tools and create deployment pipelines

	╔═══════╦═════════╗
	║ ║ ╨ ║
	╠ ╬
	╚═══════╩═════════╝